oversight

Medicare Part D: Use of Pharmacy Benefit Managers and Efforts to Manage Drug Expenditures and Utilization

Published by the Government Accountability Office on 2019-08-13.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

             United States Government Accountability Office
             Report to Congressional Requesters




             MEDICARE PART D
July 2019




             Use of Pharmacy
             Benefit Managers and
             Efforts to Manage
             Drug Expenditures
             and Utilization




GAO-19-498
                                           July 2019

                                           MEDICARE PART D
                                           Use of Pharmacy Benefit Managers and Efforts to
                                           Manage Drug Expenditures and Utilization
Highlights of GAO-19-498, a report to
congressional requesters




Why GAO Did This Study                     What GAO Found
Total expenditures for the Medicare        Medicare Part D plan sponsors used pharmacy benefit managers (PBM) to
Part D drug program exceeded $100          provide 74 percent of drug benefit management services and performed the
billion in 2016. Part D plan sponsors      remaining 26 percent of services themselves in 2016—the most recent year of
may use a PBM to provide drug benefit      data at the time of our analysis. Plan sponsors are private entities that operate
management services for Part D             drug plans; PBMs are organizations that help manage drug benefits.
coverage, such as negotiating drug
rebates and other price concessions        Rebates and other price concessions—discounts generally paid by
and paying pharmacy claims.                manufacturers to Part D plan sponsors and PBMs after the sale of a drug at the
Policymakers have sought a better
                                           pharmacy—grew faster than Part D expenditures from 2014 through 2016.
understanding of PBMs’ roles in the
                                           Specifically, gross expenditures (the amount paid to pharmacies by plan
drug supply chain and plans’ and
PBMs’ efforts to manage Part D drug
                                           sponsors, or by the PBM on the sponsor’s behalf, and by the beneficiary)
spending and use.                          increased 20 percent, to $145.1 billion. During this period, rebates and other
                                           price concessions increased 66 percent, to $29 billion—20 percent of 2016 gross
GAO was asked to examine the role of       expenditures. Consequently, net expenditures (gross expenditures less rebates
PBMs in the Part D program. This           and other price concessions) increased only 13 percent, to $116.1 billion.
report examines, among other
objectives, (1) the extent to which Part
D plan sponsors use PBMs, (2) trends       Gross Medicare Part D Expenditures, Net Part D Expenditures, and Rebates and
in rebates and other price concessions     Other Price Concessions for All Part D Drugs, 2014-2016 (in billions of dollars)
obtained by both PBMs and plan
sponsors for Part D drugs, and (3) how
PBMs earn revenue for services
provided to Part D plans.
GAO analyzed Centers for Medicare &
Medicaid Services (CMS) data on Part
D plan sponsors’ use of PBMs in 2016
as well as CMS drug expenditure,
pricing, and rebate and other price
concession data for all Part D drugs
from 2014 through 2016 (the most
recent available data at the time of our
analysis). GAO reviewed service
agreements between Part D plan
sponsors and PBMs that were
approved by CMS from January 2016          PBMs primarily earned Part D revenue through a volume-based fee paid by plan
through May 2018 and had the highest       sponsors based on PBM-processed claims; a per-member, per-month fee paid
enrollment as of June 2018. GAO            by plan sponsors; or a combination of the two. PBMs also earned revenue from
spoke with CMS officials and 38            the rebates they negotiated with manufacturers for Part D drugs, which
stakeholder groups including PBMs,         accounted for $18 billion of the $26.7 billion in rebates in 2016. PBMs retained
Part D plan sponsors, pharmacy             less than 1 percent of these rebates, passing the rest to plan sponsors. Plan
representatives and drug                   sponsors in turn may use rebates to help offset the growth in drug costs, helping
manufacturers.                             control premiums for beneficiaries.

                                           The Department of Health and Human Services provided technical comments on
                                           a draft of this report, which GAO incorporated as appropriate.
View GAO-19-498. For more information,
contact John Dicken at (202) 512-7114 or
dickenj@gao.gov.



                                                                                    United States Government Accountability Office
Contents


Letter                                                                                    1
               Background                                                                 9
               Part D Plan Sponsors Used a PBM to Provide Most Drug Benefit
                  Management Services in 2016, and Use Was Concentrated
                  among Five PBMs                                                       14
               PBMs Primarily Earned Part D Revenue through Fees Paid by
                  Plan Sponsors, Not Rebate Retention, and Reported That This
                  Differed from PBMs’ Commercial Plan Revenue                           16
               Rebates and Other Price Concessions Grew Faster Than Part D
                  Expenditures from 2014 through 2016                                   18
               Part D Drug Prices Were Significantly Lower Than List Prices for
                  Brand-Name Drugs in Retail Pharmacies; Drugs Sold in
                  Specialty Pharmacies Received Fewer Discounts                         24
               Utilization Management Was Generally Associated with Financial
                  Savings and Improved Health Indicators, but Its Effect on
                  Medication Adherence and Access Was Less Clear                        26
               Agency Comments                                                          28

Appendix I     Objectives, Scope, and Methodology                                       29



Appendix II    Medicare Part D Plan Sponsors’ Use of Pharmacy Benefit
               Managers (PBM)                                                           39



Appendix III   Information on Pharmacy Benefit Manager (PBM) Revenue
               Earned from Manufacturers and from Spread Pricing                        45



Appendix IV    Expenditures and Rebate and Other Price Concession Information
               for Medicare Part D Drugs                                                47



Appendix V     Information on Discounts Off Manufacturer List Prices for
               Brand-Name and Generic Medicare Part D Drugs                             54




               Page i                                             GAO-19-498 Medicare Part D
Appendix VI     Studies and Stakeholders’ Views on Effects of Utilization
                Management Services                                                      63



Appendix VII    Bibliography of Peer Reviewed Studies Used in GAO’s Literature
                Review                                                                   71



Appendix VIII   GAO Contact and Staff Acknowledgments                                    78


Tables
                Table 1: Rebates and Other Price Concessions and Expenditures
                        for the Highest Expenditure, Highest Utilization Brand-
                        Name Medicare Part D Drugs, 2016                                 21
                Table 2: Percentage Change in Rebates and Other Price
                        Concessions and Expenditures for the Highest
                        Expenditure, Highest Utilization Brand-Name Medicare
                        Part D Drugs, 2014-2016                                          22
                Table 3: Rebates and Other Price Concessions Received by
                        Medicare Part D Plan Sponsors That Used a Pharmacy
                        Benefit Manager (PBM) for Rebate Negotiations and
                        Those That Did Not, 2016                                         23
                Table 4: Number of Medicare Part D Plan Contracts That Used
                        Top Five Pharmacy Benefit Managers (PBM) for Drug
                        Benefit Management Services, 2016                                43
                Table 5: Expenditures and Rebates and Other Price Concessions
                        for the 444 Highest Expenditure, Highest Utilization
                        Medicare Part D Brand-Name Drugs, by Therapeutic
                        Class, 2016                                                      51
                Table 6: Gross and Net Expenditure Information for the 17
                        Medicare Part D Plan Sponsors GAO Interviewed, 2016              52
                Table 7: Medicare Part D Median Unit Drug Part D Prices as a
                        Percentage of Manufacturer List Prices for the Highest
                        Expenditure, Highest Utilization Brand-Name Drugs Sold
                        in Retail and Specialty Pharmacies, 2016                         55
                Table 8: Medicare Part D Median Unit Drug Prices as a
                        Percentage of Manufacturer List Prices by Therapeutic
                        Class for the Highest Expenditure, Highest Utilization
                        Brand-name Drugs Sold in Retail Pharmacies, 2016                 56



                Page ii                                            GAO-19-498 Medicare Part D
          Table 9: Medicare Part D Median Unit Drug Prices as a
                  Percentage of Manufacturer List Prices for the Highest
                  Expenditure, Highest Utilization Generic Drugs Sold in
                  Retail and Specialty Pharmacies, 2016                           60
          Table 10: Medicare Part D Median Unit Drug Prices as a
                  Percentage of Manufacturer List Prices by Therapeutic
                  Class for the Highest Expenditure, Highest Utilization
                  Generic Drugs Sold in Retail Pharmacies, 2016                   61
          Table 11: Medicare Part D Median Unit Drug Prices as a
                  Percentage of Manufacturer List Prices by Therapeutic
                  Class for the Highest Expenditure, Highest Utilization
                  Generic Drugs Sold in Specialty Pharmacies, 2016                62

Figures
          Figure 1: Example of the Flow of Funds and Prescription Drugs
                   through the Supply Chain when a Medicare Part D
                   Beneficiary Purchases a Drug through a Part D Plan
                   Sponsor Using a Pharmacy Benefit Manager (PBM)                 10
          Figure 2: Gross Medicare Part D Expenditures, Net Part D
                   Expenditures, Rebates and Other Price Concessions for
                   All Part D Drugs, 2014-2016 (in billions of dollars)           19
          Figure 3: Medicare Part D Median Drug Prices as a Percentage of
                   Manufacturer List Prices for Highest Expenditure and
                   Highest Utilization Brand-Name Drugs, 2016                     25
          Figure 4: Medicare Part D Plan Sponsor Contract Use of
                   Pharmacy Benefit Managers (PBM) for 10 Drug Benefit
                   Management Services, by Percent of Contracts, 2016             40
          Figure 5: Number of Pharmacy Benefit Managers (PBM) Used for
                   10 Drug Benefit Management Services by Percent of All
                   Medicare Part D Plan Sponsor Contracts, 2016                   41
          Figure 6: Pharmacy-related Price Concessions, 2014-2016                 48
          Figure 7: Rebates and Other Price Concessions Received for the
                   Highest-Expenditure, Highest-Utilization Medicare Part D
                   Brand-Name Drugs as a Proportion of Their Gross
                   Expenditures, 2016                                             50
          Figure 8: Medicare Part D Drug Prices as a Percentage of
                   Manufacturer List Prices for the Highest Expenditure,
                   Highest Utilization Generic Drugs Sold in Retail and
                   Specialty Pharmacies, 2016                                     58




          Page iii                                          GAO-19-498 Medicare Part D
Abbreviations

AWP               average wholesale price
CMS               Centers for Medicare & Medicaid Services
DIR               direct and indirect remuneration
HHS               Department of Health and Human Services
HPMS              Health Plan Management System
ISDR              Ingredient, strength, dosage, and route of administration
NADAC             National Average Drug Acquisition Cost
PACE              Program of All-inclusive Care for the Elderly
PBM               Pharmacy Benefit Manager
PDE               prescription drug event
PDP               prescription drug plan
PSAO              Pharmacy Services Administrative Organization




This is a work of the U.S. government and is not subject to copyright protection in the
United States. The published product may be reproduced and distributed in its entirety
without further permission from GAO. However, because this work may contain
copyrighted images or other material, permission from the copyright holder may be
necessary if you wish to reproduce this material separately.




Page iv                                                        GAO-19-498 Medicare Part D
                       Letter




441 G St. N.W.
Washington, DC 20548




                       July 15, 2019

                       The Honorable Susan Collins
                       Chairman
                       Special Committee on Aging
                       United States Senate

                       The Honorable Richard Neal
                       Chairman
                       Committee on Ways and Means
                       House of Representatives

                       Medicare Part D is the voluntary program that provides outpatient
                       prescription drug coverage for Medicare beneficiaries who enroll in Part D
                       drug plans. 1 Total Part D program expenditures were more than $100
                       billion in 2016, the most recent data at the time of our analysis. These
                       expenditures account for the amount paid to pharmacies by Part D plan
                       sponsors, or by a pharmacy benefit manager PBM (PBM) on the
                       sponsors’ behalf, and by beneficiaries for Part D drugs. Part D plan
                       sponsors—which are private companies—contract with the Centers for
                       Medicare & Medicaid Services (CMS) to provide this prescription drug
                       coverage to Medicare beneficiaries. 2 Plan sponsors may have multiple
                       contracts with CMS, with each contract providing one or more distinct
                       drug plans. 3 Plans may charge different monthly premiums and have
                       different beneficiary cost-sharing arrangements—such as deductibles and
                       cost-sharing for covered drugs. 4


                       1
                        Beneficiaries may receive Part D coverage through either stand-alone Part D prescription
                       drug plans (PDPs) that supplement traditional Medicare or through Medicare Advantage
                       (Part C) plans that generally must cover all Medicare benefits and usually offer Part D
                       coverage.
                       2
                        PBMs are organizations that help manage drug benefits.
                       3
                        For example, a Part D plan sponsor may have contracts covering various regions of the
                       country. Plans covered under the same contract may differ in their benefit structure, the
                       drugs they cover, and the pharmacies they contract with to fill prescriptions. In this report,
                       we refer to a drug plan or plans covered under each sponsor contract with CMS as a “Part
                       D plan sponsor contract.”
                       4
                        A deductible is a fixed dollar amount that beneficiaries must pay before coverage takes
                       effect. Part D plan sponsor payments to pharmacies generally include a portion that the
                       beneficiary pays, known as cost-sharing, which may be a flat amount (copayment) or a
                       percentage of the drug’s costs (coinsurance).




                       Page 1                                                           GAO-19-498 Medicare Part D
There are a number of services associated with providing a drug benefit,
including establishing networks of pharmacies and negotiating rebates
and other price concessions from manufacturers. One drug benefit
management service that is often performed is utilization management, a
process to help ensure that the use of drugs and other medical services is
based on medical necessity, efficiency, and appropriateness. Part D plan
sponsors may perform these drug benefit services themselves or have
them performed by PBMs. PBMs have come under scrutiny as
policymakers have attempted to better understand their role in the drug
supply chain and plan sponsors’ and PBMs’ efforts to manage Part D
drug spending and use.

You asked us to provide an overview of the role of PBMs in the Medicare
Part D program. This report examines:

1. the extent to which Part D plan sponsors use PBMs to deliver drug
   benefit management services to Medicare beneficiaries;
2. how PBMs earn revenue from the services they provide to Part D plan
   sponsors;
3. trends in rebates and other price concessions obtained by Part D plan
   sponsors and PBMs from drug manufacturers and others;
4. the extent to which prices for Part D drugs are discounted off of
   manufacturer list prices; and
5. what is known about savings and other effects of utilization
   management services commonly used in Part D plans.

To examine the extent to which Part D plan sponsors contract with PBMs
to deliver drug benefit management services to Medicare beneficiaries,
we analyzed CMS Health Plan Management System (HPMS) data for
2016, the most recent available expenditure and rebate and other price
concession data at the time of our analysis. The data identified the entity
or entities responsible for performing each of 10 drug benefit
management services under plan sponsors’ Part D contracts. 5 According
to CMS, these are the key drug benefit management services associated
with providing Part D drug coverage and include paying pharmacy claims


5
 HPMS is CMS’s web-based system through which Part D plan sponsors report their bids
and other contract information to CMS. According to CMS, Part D plan sponsors are to
report the entities responsible for performing these services on an ongoing basis.




Page 2                                                    GAO-19-498 Medicare Part D
and negotiating rebates and other price concessions. 6 CMS provided
HPMS data for the 624 Part D plan sponsor contracts that were effective
in 2016. 7 For each contract, we used the HPMS data to determine the
extent to which a plan sponsor performed a service itself, contracted with
a PBM to perform the service, or performed the service in coordination
with a PBM. In this report, we refer to any organization (other than the
plan sponsor itself) that provides one of the 10 drug benefit management
services to a plan sponsor as a PBM.

To examine how PBMs earn revenue for the services they provide to Part
D plan sponsors, we examined 20 service agreements between PBMs
and Part D plan sponsors. 8 These agreements generally contain detailed
information on the services that the PBM will provide, how the plan
sponsor will pay the PBMs for those services, and the rates pharmacies
will be paid for prescription drugs. 9 The 20 service agreements were
those approved between January 2016 and May 2018 that had the
highest enrollment in June 2018, the most recent data available at the
time of our analysis. 10

We also examined PBM revenue reported to CMS by Part D plan
sponsors in their rebates and other price concession data reports—also
referred to as direct and indirect remuneration (DIR)—in 2016, the most

6
 The eight other services are: pharmacy network development, enrollment processing,
enrollee appeals and grievance process management, customer service, management of
a pharmacy and therapeutics committee, coordination with other drug benefit programs,
pharmacy technical assistance, and drug benefit administration. Instead of being reported
as one of the 10 distinct services, utilization management services may be included under
several of the 10 prescription drug benefit services that plan sponsors report to CMS in
the HPMS data.
7
 These 624 contracts provided 4,663 unique plans and were operated by 207 Part D plan
sponsors in June 2016.
8
 The service agreements and the provisions in those agreements that we examined are
not generalizable to all service agreements that are in effect.
9
 One service agreement did not have complete information on the primary way that the
PBMs would be paid. CMS told us that payment information between Part D plan
sponsors and PBMs may be omitted from information submitted to the agency if the
sponsors consider it proprietary.
10
  Part D plan sponsors are required to provide CMS their service agreements with PBMs
for initial approval by the agency. They are also required to provide CMS with any
changes to substantive provisions of the service agreements, such as payment provisions,
but do not have to provide CMS with a new service agreement containing the changes.
See 42 U.S.C. § 1395w-115(f)(1).




Page 3                                                        GAO-19-498 Medicare Part D
recent data available at the time of our analysis. These data include
information on any price concessions made after a drug is purchased
from the pharmacy by a beneficiary. One type of price concession is a
rebate, which is generally a discount paid by drug manufacturers to a Part
D plan sponsor, or by a PBM on the sponsor’s behalf, after a beneficiary
purchases a drug. 11 These discounts may be offered in exchange for
better placement on a plan sponsor’s list of covered drugs, known as a
formulary, which encourages the use of the manufacturer’s drugs by
assigning them to tiers within the formulary that have lower beneficiary
cost sharing. 12 Plan sponsors and PBMs may receive other price
concessions that lower the price of a drug. For example, plan sponsors
may receive fees from pharmacies based on their performance, which
affect prices for certain drugs since the performance fees affect the
amount the plan sponsor pays the pharmacy. 13 The rebate and other
price concession reports also include information on any revenue earned
by PBMs through their retaining a portion of negotiated rebates. We
define the gross price of a drug as the total amount paid to the pharmacy
by the Part D plan sponsor, the PBM on the sponsor’s behalf, and the
beneficiary; gross price less rebates and other price concessions is the
net price. The rebate and other price concession reports to CMS also
include monies that are not concessions used in the calculation of net
price, such as certain sources of PBM revenue, including fees paid by
manufacturers to PBMs for certain services, as well as spread pricing—

11
   Beneficiary cost-sharing is based on the price of the drug at the point of sale. Therefore,
if a manufacturer provides a rebate after the point of sale, which, as noted, is generally the
case, the beneficiary’s cost sharing is not reduced by the amount of the rebate. Other
factors may affect the amount of beneficiary cost-sharing, such as beneficiaries paying
more for drugs purchased at pharmacies not within the Part D plan sponsor’s preferred
pharmacy network.
12
  A formulary is a Part D plan sponsor’s list of covered drugs. Plan sponsors may assign
drugs to different tiers within a formulary that correspond to different levels of beneficiary
cost sharing. The tiers with lower cost sharing correspond to more favorable placement in
the formulary to encourage beneficiaries to use lower cost drugs. Although rebates and
other price concessions do not directly affect the price beneficiaries pay at the pharmacy,
which is based on the price before any price concessions are applied, the presence of a
rebate may affect their cost-sharing in other ways, such as by placing the drug on a
preferred formulary tier with lower cost sharing in exchange for the rebate.
13
  Part D plan sponsors and PBMs may enter into performance arrangements with
pharmacies where they either pay a fee or receive a bonus based on their performance,
such as a specified percent of prescriptions dispensed for a generic drug (instead of a
brand-name drug). The revenue paid to pharmacies is considered an incentive payment
and monies received from pharmacies by plan sponsors, or PBMs on their behalf, is
considered a price concession.




Page 4                                                           GAO-19-498 Medicare Part D
where PBMs earn revenue by keeping the difference between the amount
they charged the pharmacy and the amount the they charged the plan for
a drug.

To examine trends in rebates and other price concessions obtained by
Part D plan sponsors and PBMs from manufacturers and others for Part
D drugs, we analyzed plan sponsors’ gross and net expenditures for
these drugs from 2014 through 2016. For a given drug, gross
expenditures reflect what was paid to the pharmacy by the Part D plan
sponsor, PBMs on the sponsor’s behalf, and the beneficiary. To calculate
gross expenditures, we used Medicare prescription drug event (PDE)
data to calculate gross brand-name and generic drug expenditures and
utilization for all Part D plan sponsors. 14 Net expenditures reflect any
rebates and other price concessions obtained by Part D plan sponsors
and PBMs after a beneficiary receives a drug. To calculate net
expenditures, we obtained rebate and other price concession information
and subtracted it from plan sponsors’ brand-name and generic gross
expenditures. We identified brand-name and generic drugs by grouping
expenditure claims with the same active ingredient, strength, dosage
form, and route of administration. 15 We also used CMS Part D enrollment
data to examine gross and net expenditures per Medicare beneficiary for
sponsor contracts in 2016. 16 We also examined differences in the amount




14
  Part D plan sponsors—both PDPs and Medicare Advantage drug plans—submit a PDE
record to CMS for each time a beneficiary obtains a prescription drug. The PDE record
contains information on the beneficiary receiving the drug, the price paid by the plan
sponsor to the pharmacy, and applicable beneficiary cost-sharing. We calculated
expenditures based on a drug’s ingredient cost, dispensing fees, sales tax, and applicable
vaccine administration fees for all Part D drugs. We excluded PDE claims billed under
Programs of All-Inclusive Care for the Elderly plan contracts because they are exempted
from certain Part D requirements, such as charging beneficiaries cost-sharing. We also
excluded compounded drugs, which are tailor-made by a pharmacy for a beneficiary and
over-the-counter drugs as they are generally not covered by Medicare Part D. We used
information from Red Book, a compendium published by Truven Health Analytics, to
identify brand-name and generic drugs and to determine drugs’ therapeutic class, dose,
and route of administration.
15
  The dosage form is the physical form in which a drug is produced and dispensed, such
as a tablet or capsule; route of administration is the way of administering a drug to a site in
a patient, such as taking a drug orally.
16
  We used June 2016 data to indicate enrollment, as this month has relatively stable
enrollment as it does not fall within an annual open enrollment period where beneficiaries
may change their Part D coverage.




Page 5                                                           GAO-19-498 Medicare Part D
of rebates and other price concessions relative to expenditures obtained
by Part D plan sponsors that used a PBM, relative to those that did not. 17

To obtain more information on drugs that have the greatest fiscal impact
on the Part D program and its beneficiaries, we calculated gross and net
expenditures for the brand-name and generic drugs with the highest
expenditures, highest utilization, and highest expenditure per utilization in
2016. For both brand-name and generic drugs, we identified the following:
the 200 brand-name and 200 generic drugs with the highest expenditures
in 2016; the 200 brand-name and generic drugs with the highest
utilization in 2016 (based on number of 30-day prescriptions); and the 200
brand-name and generic drugs with the highest expenditures per
utilization (i.e., highest expenditure per number of 30-day prescriptions).
As a result of overlap in the groups of drugs, these criteria yielded lists of
the 444 unique highest expenditure, highest utilization brand-name drugs
and the 476 unique highest expenditure, highest utilization generic drugs.
Together, these 920 highest expenditure, highest utilization brand-name
and generic drugs accounted for 81 percent of Part D expenditures in
2016. 18

To examine the extent to which Part D drug prices are discounted off of
manufacturer list prices, we compared the median gross and net prices
for the 444 brand-name and 476 generic highest expenditure, highest
utilization drugs to (1) list prices established by manufacturers and (2) the
cost to pharmacies of acquiring these drugs. For the list price, we used
the 2016 average wholesale price (AWP)—which we refer to as
manufacturer list price—which reflects the average price manufacturers
suggest wholesalers charge pharmacies for a drug. 19 For pharmacy
acquisition costs, which reflect the price pharmacies paid to obtain the
drug, we used retail community pharmacy acquisition cost data from




17
  We determined PBM use through 2016 HPMS data that identified whether a Part D plan
sponsor contract used a PBM for rebate and price concession negotiations with
manufacturers, pharmacies, or others.
18
  Of the 920 highest expenditure, highest utilization brand-name and generic drugs, the
444 brand-name drugs accounted for 65 percent of expenditures, and the 476 generic
drugs accounted for 16 percent of expenditures.
19
 AWP is a list price and does not reflect the actual price paid by pharmacies or
wholesalers for a drug. We obtained AWP information from Truven Analytics’ Red Book.




Page 6                                                        GAO-19-498 Medicare Part D
National Average Drug Acquisition Cost (NADAC) data. 20 Part D plan
sponsors and PBMs acting on the sponsor’s behalf may negotiate prices
paid to pharmacies that are lower than manufacturers’ list price, but
higher than pharmacies’ acquisition costs. Separately, the plan sponsor
or PBM may also receive rebates and other price concessions that are
not part of their payments to pharmacies, but are reflected in their net
price. We also calculated a gross Part D drug price using 2016 PDE data
by dividing gross per unit expenditures for a given drug by the total
quantity dispensed for the drug. 21 To calculate a net Part D price, we
subtracted rebates and other concessions per quantity dispensed from
the gross per unit price. We separated drugs sold in retail community
pharmacies from those sold in specialty pharmacies, as they dispense
low-volume and high-cost drugs to patients undergoing intensive
therapies for illnesses. 22 For each drug, we then determined median
pharmacy acquisition costs, median gross Part D prices, and median net
Part D prices as a proportion of median manufacturer list prices.

To examine what is known about savings and other effects of utilization
management services commonly used in Part D plans, we conducted a
literature search for studies that examined the effect of utilization
management services in Part D (regardless of whether they were
provided by a PBM or another entity) on the following outcomes: (1)
financial costs or savings, (2) beneficiaries’ health indicators, and (3)
beneficiaries’ access to clinically appropriate medications or taking their
medications as prescribed (adherence). The literature search was
performed from April 2018 to July 2018 using keyword searches in
bibliographic databases, including ProQuest, EBSCO, and Scopus. We
limited our search to peer-reviewed studies published beginning in




20
  NADAC is obtained from a survey of community retail pharmacies and was developed to
provide a national pricing benchmark for states’ Medicaid programs. NADAC does not
contain data from non-retail pharmacies, such as mail-order or specialty pharmacies.
21
 Quantity dispensed is the unit of measure for the drug such as milliliters or milligrams.
22
  We determined whether a drug was sold in retail or specialty pharmacies based on
NADAC data. We found that 97 percent of the 444 highest expenditure, highest utilization
brand-name drugs that lacked NADAC were listed as specialty drugs on a Part D plan
sponsor’s formulary. Given this, we refer to drugs that do not have a NADAC price as
those sold in specialty pharmacies.




Page 7                                                          GAO-19-498 Medicare Part D
2006—the year the Part D program began. 23 We identified and reviewed
52 studies that met these criteria.

For all five of our objectives, we obtained the perspectives of
stakeholders on Part D plan sponsors’ use of PBMs as well as their
perspectives on sponsors’ efforts to control Part D expenditures and drug
utilization. These stakeholders consisted of representatives from 17
small, mid-sized, and large Part D plan sponsors; seven PBMs; three
drug manufacturers; a wholesaler and pharmacy services administrative
organization; and a patient advocacy organization. 24

For all of the data we analyzed, we took steps to assure their reliability,
including interviewing knowledgeable officials, conducting data checks,
and comparing to published information when available. After taking
these steps, we determined that the data were sufficiently reliable for the
purposes of our reporting objectives. Appendix I provides additional
details on our scope and methodology.

We conducted this performance audit from May 2017 to July 2019 in
accordance with generally accepted government auditing standards.
Those standards require that we plan and perform the audit to obtain
sufficient, appropriate evidence to provide a reasonable basis for our
findings and conclusions based on our audit objectives. We believe that
the evidence obtained provides a reasonable basis for our findings based
on our audit objectives.




23
  We analyzed utilization management services provided to Medicare beneficiaries
without regard to who provided the service.
24
  The 17 Part D plan sponsors consisted of 11 that had the 20 contracts with the largest
enrollment in 2016 (enrollment in these contracts accounted for 82 percent of Part D
enrollment in 2016); three that had contracts with enrollment at or just below the median
2016 contract enrollment; and three that had three contracts with enrollment at or below
the bottom enrollment quartile. We spoke with representatives from the six PBMs that
provided the most drug benefit services to Part D plan sponsors in 2016 as identified by
our analysis of HPMS data. We also spoke with the PBM that provided the eighth most
drug benefit services to Part D plan sponsors.




Page 8                                                         GAO-19-498 Medicare Part D
Background
Prescription Drug Supply   Several entities are involved with, and pay different prices for, prescription
Chain                      drugs as they move from the manufacturer to the beneficiary (a system
                           referred to as the prescription drug supply chain). In general,
                           manufacturers develop and sell their drugs to wholesalers, and
                           wholesalers then sell the drugs to pharmacies. In the Part D program,
                           CMS pays Part D plan sponsors to provide drug coverage, and plan
                           sponsors may charge beneficiaries monthly premiums in exchange for
                           coverage. Plan sponsors and PBMs negotiate reimbursement rates for
                           the drugs provided to beneficiaries. When the beneficiary purchases a
                           drug, the pharmacy is paid by the Part D plan sponsor, or through the
                           PBM on the sponsor’s behalf, and by the beneficiary through any
                           applicable cost-sharing. (See fig. 1 for a flow chart showing the
                           relationship between certain entities in the prescription drug supply chain
                           when a Part D plan sponsor uses a PBM.)




                           Page 9                                                GAO-19-498 Medicare Part D
Figure 1: Example of the Flow of Funds and Prescription Drugs through the Supply Chain when a Medicare Part D Beneficiary
Purchases a Drug through a Part D Plan Sponsor Using a Pharmacy Benefit Manager (PBM)




                                        Note: This flow chart illustrates the use of a PBM by a Medicare Part D plan sponsor to pay
                                        prescription drug claims and negotiate rebates and other price concessions with manufacturers and
                                        others. In some cases, Part D plan sponsors perform these activities themselves or in conjunction
                                        with a PBM. For example, a plan sponsor may use a PBM to pay the pharmacy for a drug, while
                                        negotiating rebates directly with the manufacturer. In addition, other funds flow through the
                                        prescription drug supply chain. For example, the Centers for Medicare & Medicaid Services (CMS)
                                        makes prospective monthly payments to Part D plan sponsors based on plan sponsors’ estimates of
                                        providing drug coverage to beneficiaries. These payments are reconciled by CMS at the end of the
                                        year to ensure that payments reflect actual drug costs minus rebates and other price concessions.
                                        Beneficiaries may also pay plan sponsors monthly premiums in exchange for their drug coverage.




                                        Page 10                                                             GAO-19-498 Medicare Part D
Prescription Drug Plan   Services associated with developing and managing a prescription drug
Services                 plan performed by PBMs, Part D plan sponsors, or both, include:

                         •     Formulary development. Determining the list of drugs covered under
                               the plan (the formulary), including assignment of covered drugs to
                               tiers that correspond to different levels of beneficiary cost sharing and
                               placing restrictions on drugs included in the formulary. Part D plan
                               sponsors submit formularies for their plans to CMS for review and
                               approval annually.
                         •     Pharmacy network development. Creating a network of pharmacies
                               where beneficiaries may fill their prescriptions and negotiating drug
                               prices and reimbursement rates with those pharmacies. This can also
                               include developing “preferred networks,” whereby beneficiaries pay
                               lower cost-sharing and pharmacies agree to receive lower prices for
                               drugs in exchange for increased volume of prescriptions purchased. 25
                         •     Utilization management services. Utilization management services
                               include processes such as:
                               •   Prior authorization. A requirement that beneficiaries obtain
                                   approval for a drug by the PBM or plan sponsor before obtaining
                                   the drug if it is to be covered by the plan.
                               •   Step therapy. A requirement where more expensive drugs are
                                   covered only if beneficiaries try less expensive alternatives first
                                   and find them not to be effective.
                               •   Medication therapy management. A program required by CMS
                                   designed to improve medication adherence and reduce the risk of
                                   adverse drug events through discussion with targeted
                                   beneficiaries and prescriber intervention. 26
                               •   Drug utilization review. A concurrent examination by the PBM or
                                   plan sponsor of prescriptions at the time of purchase by the
                                   beneficiary to assess safety considerations, such as potential
                                   adverse interactions, and compliance with clinical guidelines
                                   (including quantity and dose). These reviews can also occur

                         25
                             See 42 C.F.R. § 423.100 (2018).
                         26
                            Beneficiaries are automatically eligible for a Part D plan sponsor’s CMS-mandated
                         medication therapy management program if they meet they meet the following criteria: (1)
                         have at least two chronic diseases; (2) are taking at least two Part D drugs; and (3) are
                         likely to incur annual costs for covered Part D drugs greater than or equal to a specified
                         cost threshold. Plan sponsors may include additional beneficiaries in their program,
                         independent of these criteria.




                         Page 11                                                       GAO-19-498 Medicare Part D
                                retrospectively to analyze beneficiaries’ drug utilization and
                                physicians’ prescribing patterns.
                      •   Negotiation of rebates from manufacturers. Negotiating rebates for
                          Part D plan sponsors with manufacturers in exchange for driving more
                          utilization of a manufacturer’s drug. This can include more favorable
                          placement on the sponsor’s formulary. The rebate terms do not have
                          to be disclosed to the public, but plan sponsors must report rebate
                          amounts to CMS.

PBM Revenue           PBMs may earn revenue from providing drug benefit management
                      services to Part D plan sponsors in a number of ways, including: (1)
                      payments from plan sponsors for administering services, such as drug
                      benefits claim processing; (2) retention of a portion of drug rebates that
                      PBMs negotiate on behalf of the plan sponsor and fees for managing and
                      distributing those rebates; (3) spread pricing; and (4) payments from
                      manufacturers for various services. PBMs may provide drug benefit
                      management services to Part D plan sponsors and commercial plans,
                      such as employer-sponsored health plans. Commercial plans may pay
                      PBMs in ways similar to Part D plans (e.g., rebate retention and claims
                      processing fees).


Part D Coverage and   Part D plan sponsors are also required to provide access to all or
Payments              substantially all drugs covered under certain therapeutic classes of drugs,
                      known as Medicare protected classes: (1) anticonvulsants, (2)
                      antidepressants, (3) antineoplastics, (4) antipsychotics, (5) antiretrovirals,
                      and (6) immunosuppressants for the treatment of transplant. Plans are
                      limited in the formulary restrictions they can apply to these drugs.
                      Additionally, CMS generally requires Part D plan sponsors to provide
                      coverage for at least two drugs in each class.

                      CMS makes payments prospectively to Part D plan sponsors for
                      beneficiary drug coverage. CMS pays plan sponsors monthly, and these
                      payments are determined through annual bids submitted in June of the
                      preceding program year, which runs from January 1 through December
                      31. Those bids reflect the plan sponsors’ estimates of program costs and
                      rebates and other price concessions that the sponsor expects to receive
                      during the ensuing program year. At the end of the program year, CMS
                      reviews cost data submitted by plan sponsors through PDE records and
                      their submission of rebate and other price concession data and compares
                      estimated payments with actual costs incurred, with CMS either
                      reclaiming some funds or making additional payments. Thus, the final



                      Page 12                                                 GAO-19-498 Medicare Part D
                          plan payments by CMS are based on the costs actually incurred by Part
                          D plan sponsors minus rebates and other price concessions that are
                          either passed along to the plan sponsors or retained by the PBMs.


Implications of Rebates   Rebates and other price concessions reduce the cost of the Part D
and Other Price           program to beneficiaries and the federal government. In developing their
                          bids, Part D plan sponsors may subtract rebates and other price
Concessions
                          concessions that are passed along to them from their estimated drug
                          costs. 27 When they do, rebates and other price concessions reduce a
                          plan sponsor’s estimate of liability that is reflected in bid amounts, which,
                          in turn, reduce beneficiary premiums because they are based, in part, on
                          the bid amount. This downward pressure on premiums is one reason that
                          premiums remained relatively unchanged between 2010 and 2015,
                          according to CMS, even though total gross Part D drug costs grew about
                          12 percent per year in that period.

                          Rebates have additional implications for Part D beneficiaries and the Part
                          D program more generally. Since beneficiary cost sharing is calculated
                          based on the price of the drug at the time of purchase (i.e., before rebates
                          are paid), beneficiaries pay higher cost sharing than they would if rebates
                          were paid at the point of sale. In addition, higher pre-rebate drug prices
                          may result in beneficiaries more quickly reaching the catastrophic
                          coverage phase, where the federal government’s share of drug costs
                          increases, and the plan sponsors’ share decreases. 28


                          27
                            Part D plan sponsor bids may be based on other factors, including estimated drug costs,
                          beneficiary cost sharing, administrative expenses, and profit.
                          28
                            There are multiple phases in the Part D program that beneficiaries may pass through
                          based on their cost sharing (which do not include premiums). For 2019, after a $415
                          deductible is met, beneficiaries enter the initial coverage period, where enrollees cover 25
                          percent of costs and plans cover 75 percent of costs until drug costs exceed $3,820, when
                          enrollees reach the coverage gap. In the coverage gap (“donut hole”) phase, brand-name
                          drug costs are covered by drug manufacturers (70 percent drug discounts), plan sponsors
                          (5 percent), and beneficiaries (25 percent). For generic drugs, plan sponsors cover 63
                          percent of costs, and beneficiaries cover 37 percent. In 2019, beneficiaries reach the
                          catastrophic phase when their cost sharing reaches $5,100. The federal government then
                          covers 80 percent of costs, the plan sponsor covers 15 percent, and the beneficiary
                          covers 5 percent. The coverage gap will be “closed” beginning in 2020, with the
                          beneficiary covering 25 percent of costs for both brand-name and generic drugs. The
                          federal government pays subsidies to Part D plan sponsors that cover about 75 percent of
                          the plan premium and is also responsible for payment 80 percent of costs in the
                          catastrophic phase.




                          Page 13                                                        GAO-19-498 Medicare Part D
                        Seventy-four percent of the drug benefit management services provided
Part D Plan Sponsors    under 624 Part D plan sponsor contracts were performed by a PBM alone
Used a PBM to           or in conjunction with a Part D plan sponsor in 2016. 29 We found that plan
                        sponsors performed the remaining 26 percent of services themselves. In
Provide Most Drug       addition, a PBM was used to provide one or more of the 10 key drug
Benefit Management      benefit management services under nearly all of the 624 Part D plan
                        sponsor contracts (99.7 percent), and the manner in which they used
Services in 2016, and   them varied, as summarized below: 30
Use Was
                            Number of drug benefit management services provided. Part D
Concentrated among      •
                            plan sponsor contracts varied by the number of services provided by
Five PBMs                   PBMs. Eighty-nine percent of Part D plan sponsor contracts used a
                            PBM alone or in conjunction with a plan sponsor for at least half of the
                            10 drug benefit management services; 15 percent of contracts used a
                            PBM alone or with a plan sponsor for all 10 services.
                        •   Number of PBMs used. Part D plan sponsor contracts varied in the
                            number of PBMs used to provide one or more of the 10 drug benefit
                            management services. Fifty-four percent of contracts used one PBM,
                            35 percent used two or three PBMs, and 11 percent used four or more
                            PBMs.
                        •   Types of drug benefit management services provided. Part D plan
                            sponsor contracts varied by the drug benefit management services
                            they used a PBM to provide. PBMs alone or with the plan sponsor
                            more frequently provided claims adjudication (99 percent of Part D
                            plan sponsor contracts), pharmacy network development (92 percent),
                            and rebate and other price concession negotiations (83 percent). In
                            contrast, PBMs alone or with the plan sponsor less frequently
                            provided a pharmacy and therapeutics committee (45 percent),

                        29
                          We determined the proportion of the 10 drug benefit management services in the HPMS
                        database that was provided by a PBM to the 624 Part D plan sponsor contracts included
                        in our analysis. PBMs were used to perform 4,592 of the 6,240 drug benefit management
                        services (74 percent) provided to the 624 contracts. We counted plan sponsor contracts
                        as using a PBM to provide a drug benefit management service if the PBM provided the
                        service alone or in combination with the plan. We found that that PBMs and Part D plan
                        sponsors together performed 20 percent of drug benefit management services, and PBMs
                        alone performed 54 percent of drug benefit management services.
                        30
                          Some of the entities used by Part D plan sponsors to provide drug benefit management
                        services may have other lines of business besides PBM services. For example, Xerox,
                        which provides print and digital document services, provided drug benefit management
                        services such as beneficiary customer services in conjunction with a plan sponsor or PBM
                        to 66 plan sponsor contracts.




                        Page 14                                                      GAO-19-498 Medicare Part D
     enrollee appeals and grievance process-management (30 percent),
     and enrollment processing (34 percent).

Part D plan sponsors mainly used five PBMs in 2016. Of the 103 PBMs
that provided at least one drug benefit management service to the 624
Part D plan sponsor contracts in 2016, the following five provided at least
one service to 528 (85 percent) plan sponsor contracts in 2016: CVS
Caremark, OptumRx, Express Scripts, Medimpact, and Argus. 31 These
five PBMs also provided the largest number of services to Part D plan
contracts in 2016. 32 For example, CVS Caremark, by itself or with another
PBM or plan sponsor, provided 17 percent of services that PBMs
provided to Part D plan sponsors’ contracts in 2016, the most of any
PBM. 33

See appendix II for more information on variation in Part D plan sponsor
contracts’ use of PBMs, factors that influence sponsors’ decision to use a
PBM, and additional information on the PBMs used by Part D plan
sponsors.




31
  For example, CVS Caremark provided at least one service to 25 percent of contracts,
which accounted for 30 percent of total Part D enrollment.
32
  The share of the PBMs’ total business that Medicare represents varied widely among
the top five PBMs. Specifically, the PBMs told us that Part D accounted for between 1
percent and 57 percent of the total number of individuals for whom they provide drug
coverage. In addition, the five largest PBMs in the Part D program included in our analysis
may vary from sources that do not focus solely only on the Medicare Part D PBM market,
or that used a different metric of market share than number of drug benefit management
services (e.g., number of prescription claims adjudicated by the PBM).
33
  The remaining four PBMs—OptumRx, Express Scripts, MedImpact, and Argus—
provided 14 percent, 10 percent, 9 percent, and 5 percent of the drug benefit management
services respectively. A Part D plan sponsor may use more than one PBM to provide a
drug benefit management service. In this instance, we counted each respective PBM as
providing the same service in our summary counts. Argus is known as DST Pharmacy
Solutions as of September 2017.




Page 15                                                        GAO-19-498 Medicare Part D
                     Our review of 20 service agreements between Part D plan sponsors and
PBMs Primarily       PBMs found that the primary revenue source for PBMs from services they
Earned Part D        provided to Part D plans was (1) a volume-based fee paid by plan
                     sponsors based on the number of paid claims that the PBM processed;
Revenue through      (2) a flat monthly per-member, per-month fee paid by plan sponsors; or
Fees Paid by Plan    (3) a combination of the two. Nineteen of the 20 service agreements that
                     we reviewed stated that PBMs were to be paid in one of these ways. 34
Sponsors, Not        None of the service agreements tied these fees to the price of a drug paid
Rebate Retention,    to the pharmacy. 35 Representatives we interviewed from all seven of the
                     PBMs confirmed that a Part D plan sponsor-paid fee for the PBM’s
and Reported That    services was the primary way they earned revenue from their Part D
This Differed from   clients.
PBMs’ Commercial     We also examined PBM revenue reported to CMS by Part D plan
Plan Revenue         sponsors in their rebates and other price concession data—also referred
                     to as direct and indirect remuneration (DIR)—in 2016, the most recent
                     data available at the time of our analysis. These data show that PBMs
                     passed nearly all rebates received from manufacturers through to Part D
                     plan sponsors in 2016. Part D plan sponsors reported to CMS that, of the
                     approximately $18 billion in rebates that PBMs negotiated with
                     pharmaceutical manufacturers that year, PBMs retained $74.3 million, or
                     about 0.4 percent, and passed through the remaining 99.6 percent to plan
                     sponsors. 36

                     The small amount of PBM rebate retention in the Part D program was
                     also reflected in the service agreements we examined and in our
                     interviews with PBM representatives. Sixteen of the 20 service
                     agreements that we reviewed included provisions that required the PBM
                     to pass through all rebates to the Part D plan sponsor; one other
                     agreement required at least 95 percent to be passed through to the plan

                     34
                       The provisions in the remaining service agreement that we reviewed did not include
                     such specific information on how the PBM was to be paid.
                     35
                       While some service agreements detailed the amount of the fees for the various services
                     that the PBM would provide, it is not possible to tell from the agreements how much
                     revenue those services generated for the PBM because no agreement provided the
                     volume of business for any service.
                     36
                       PBMs negotiated $18 billion of the $26.7 billion in rebates that were negotiated on
                     behalf of Part D plan sponsors. Plan sponsors negotiated $7.3 billion in rebates. We were
                     unable to determine the entity responsible for negotiating the remaining $1.3 billion in
                     rebates as we could not determine whether the PBM or plan sponsor negotiated rebates
                     for certain Part D plan sponsor contracts.




                     Page 16                                                       GAO-19-498 Medicare Part D
sponsor. The other three service agreements that we reviewed either did
not include provisions related to rebate retention or redacted such
information. Officials we interviewed from four of the seven PBMs told us
their PBMs passed through to Part D plan sponsors all rebates obtained
from manufacturers. Representatives of one PBM noted that plan
sponsors, in turn, may use rebates to help offset the growth in drug costs,
helping lower premiums for beneficiaries. Representatives from the other
three PBMs noted that the amount of retained rebates was relatively
small, consistent with the data reported to CMS.

PBMs and Part D plan sponsors may earn non-rebate revenue from
manufacturers for providing certain services. The service agreements we
examined included examples of this revenue, including fees for rebate
program administration, prescriber education programs, and programs
designed to ensure patients adhere to, and comply with,
recommendations regarding a particular prescription. 37 The full amount
that PBMs and Part D plan sponsors earned from manufacturers for non-
rebate services in 2016 was $516.5 million. Although CMS requires these
fees to be reported to the agency by plan sponsors, CMS does not break
out how much of the money was received by PBMs and how much was
received by plan sponsors.

PBMs earned little Part D revenue from spread pricing—keeping the
difference between the amount the PBM paid the pharmacy for a drug
and the amount the PBM charged the plan for the drug, from 2014
through 2016. PBMs earned about $300,000 from spread pricing in 2016,
according to CMS rebate and other price concession data. CMS data also
show that PBMs earned no revenue from spread pricing in either 2014 or
2015. PBMs generally earn more from spread pricing and rebate retention
from commercial plans than they do from Part D, according to officials
from three PBMs. Officials from two of these PBMs said CMS reporting
requirements have removed much of the incentive in Part D for PBMs to
earn revenue from spread pricing because of the complexity of the
requirements and the criticism from health care providers when reports to
CMS containing these amounts are publicized.

See appendix III for more information on Part D plan sponsor reporting to
CMS of the amounts of revenue—other than rebates and discounts—that

37
  We were unable to determine the amount of revenue that PBMs obtained from these
services because the volume of usage of services was not contained in any service
agreements.




Page 17                                                   GAO-19-498 Medicare Part D
                      manufacturers provide to their PBMs; and on PBM and Part D plan
                      sponsor perspectives on PBM revenue earned from spread pricing, the
                      effect of CMS requirements on spread pricing revenue, and differences
                      between PBMs’ Part D and commercial business lines.


                      Growth in the amount of rebates and other price concessions provided by
Rebates and Other     manufacturers and others to Part D plan sponsors and PBMs outpaced
Price Concessions     growth in gross and net Part D expenditures for all brand-name and
                      generic drugs from 2014 through 2016. Gross expenditures reflect what
Grew Faster Than      was paid to the pharmacy by the Part D plan sponsor—or the PBM on the
Part D Expenditures   sponsor’s behalf—and by the beneficiary for a given drug. Net
                      expenditures reflect any rebates and discounts obtained by plan sponsors
from 2014 through     and PBMs after a beneficiary receives a drug. During this time, gross Part
2016                  D expenditures increased 20 percent, from $120.7 billion in 2014 to
                      $145.1 billion in 2016. The amount of rebates and other price
                      concessions obtained for these drugs increased 66 percent during the
                      same period, from $17.5 billion to $29 billion. As a result, rebates and
                      other price concessions as a proportion of gross expenditures increased
                      from 14 percent of gross expenditures in 2014 to 20 percent in 2016. This
                      resulted in an increase in net Part D expenditures of 13 percent, from
                      $103.2 billion in 2014 to $116.1 billion in 2016 (see fig. 2).




                      Page 18                                            GAO-19-498 Medicare Part D
Figure 2: Gross Medicare Part D Expenditures, Net Part D Expenditures, Rebates
and Other Price Concessions for All Part D Drugs, 2014-2016 (in billions of dollars)




Notes: We used CMS data to analyze brand-name and generic drug expenditures for Part D
contracts from 2014 through 2016. We excluded expenditures from contracts that participated in the
Medicare Program of All-Inclusive Care because they are exempted from Part D requirements, such
as charging beneficiaries cost-sharing. We also excluded expenditures for compounded drugs which
are tailor-made by a pharmacy for a beneficiary and over-the-counter drugs as they are generally not
covered by Medicare Part D.
Gross expenditures reflect what was paid to the pharmacy by the Part D plan sponsor, pharmacy
benefit managers on the sponsor’s behalf, and the beneficiary for a given drug. Net expenditures
reflect any rebates and other price concessions obtained by plan sponsors and PBMs after a
beneficiary receives a drug as reported in the direct and indirect remuneration data.




Rebates accounted for most of the total of rebates and other price
concessions obtained for Part D drugs from 2014 through 2016. Rebates
are generally paid by manufacturers to Part D plan sponsors, or PBMs on
sponsors’ behalf, after a drug is purchased from a pharmacy. In 2016,
rebates accounted for 92 percent ($27 billion) of the $29.1 billion in
rebates and other price concessions. The proportion was generally
consistent in 2014 and 2015, with rebates accounting for 93 and 91
percent of total rebates and other price concessions, respectively.


Page 19                                                               GAO-19-498 Medicare Part D
Pharmacy-related price concessions, which include any monies obtained
by plan sponsors and PBMs from a pharmacy after a beneficiary
purchases a drug, accounted for nearly all the rest of rebates and other
price concessions—7 percent—in 2016. 38 The amount of pharmacy-
related price concessions increased 295 percent from 2014 through 2016
($538 million to $2.1 billion).

The 444 highest expenditure, highest utilization brand-name drugs
accounted for the majority of expenditures and received the vast majority
of rebates and other price concessions in 2016. These drugs accounted
for 65 percent of the $145 billion in Part D expenditures and received 90
percent of the $29.1 billion in rebates and other price concessions
obtained for Part D drugs. 39 Of the 444 highest expenditure, highest
utilization brand-name drugs in 2016, the 200 highest utilization and the
200 highest expenditure drugs received a greater amount of rebates and
other price concessions than the 200 highest expenditure per utilization
drugs. 40 (See table 1.) Furthermore we found that brand-name drugs
received greater amounts of rebates and other price concessions than
generic drugs. 41 Specifically, among the 444 highest expenditure, highest
38
  According to CMS, pharmacy-related price concessions includes payments to and from
pharmacies that affect the net price of the drug purchased by the Part D plan sponsor
after the point-of-sale. These include reconciling any differences between the contracted
payment rate by the PBM or Part D plan sponsor to the pharmacy and the rate paid to the
pharmacy at the point-of-sale. They also include bonuses paid by the plan sponsor or
PBM to the pharmacy or fees paid by the pharmacy to the plan sponsor or PBM based on
how well the pharmacy met certain agreed upon performance metrics, such as the generic
dispensing rate. In 2016, payments from pharmacies to plan sponsors totaled $2.3 billion
and payments from plan sponsors to pharmacies totaled $211 million. The net of these
payments—$2.1 billion—is reported as pharmacy-related price concessions.
39
  We analyzed Part D expenditure and rebate and other price concessions data from the
prescription drug event and DIR data sets for three groups of brand-name drugs: the 200
drugs with the highest expenditures, the 200 drugs with the highest utilization, and the 200
drugs with the highest expenditures per utilization. This resulted in a group of 444 unique
brand-name drugs across the three groups. We focused our analysis on brand-name
drugs, as they received the majority of rebates and other price concessions in 2016.
40
  There was a strong correlation between the amount of rebates and other price
concessions with expenditures (correlation coefficient of 0.91) and utilization (correlation
coefficient of 0.74) for the 444 highest expenditure, highest utilization brand-name drugs.
The correlation coefficient is a statistical measure of association, ranging in value from
negative 1 to positive 1, with negative 1 indicating a perfect negative correlation, 0 an
absence of correlation, and positive 1 a perfect positive correlation.
41
  Rebates are used more frequently for drugs where there is competition and for which
there are therapeutic substitutes and are rarely available for generic drugs. See G.
Dieguez, M. Alston, and S. Tomicki, A Primer on Prescription Drug Rebates: Insights Into
Why Rebates Are A Target For Reducing Prices (Milliman, Inc. May 2018).




Page 20                                                          GAO-19-498 Medicare Part D
                                                               utilization brand-name drugs and the 476 highest expenditure, highest
                                                               utilization generic drugs, brand-name drugs received 98 percent of
                                                               rebates and other price concessions in 2016.

Table 1: Rebates and Other Price Concessions and Expenditures for the Highest Expenditure, Highest Utilization Brand-Name
Medicare Part D Drugs, 2016

 Brand-Name Drug                                 Gross Part D              Rebates and other         Net expenditures           Rebates and       Median number
 category                                        expenditures              price concessions       (billions of dollars)         other price      of beneficiaries
                                          (billions of dollars)           (billions of dollars)                             concessions as         receiving each
                                                                                                                             a proportion of                 drug
                                                                                                                                      gross
                                                                                                                               expenditures
                                                                                                                                   (percent)
 200 Highest-expenditure                                        85.3                        24.6                   60.7                     29               52,847
 200 Highest-utilization                                        62.7                        22.6                   40.1                     36               95,852
 200 Highest-expenditure                                        20.3                         2.3                   17.9                     12                   301
 per utilization
Source: GAO analysis of Centers for Medicare & Medicaid Services (CMS) data. I GAO-19-498

                                                               Note: We analyzed CMS Part D expenditure and rebate and other price concession data from the
                                                               prescription drug event and direct and indirect remuneration data sets for three groups of brand-name
                                                               drugs: the 200 drugs with the highest expenditures, the 200 drugs with the highest utilization, and the
                                                               200 drugs with the highest expenditures per utilization. We identified drugs that had common
                                                               ingredients, strengths, dose, and route of administration and combined them. This resulted in a group
                                                               of 444 unique brand-name drugs across the three groups.
                                                               Gross expenditures reflect what was paid to the pharmacy by the Part D plan sponsor, pharmacy
                                                               benefit managers on the sponsor’s behalf, and the beneficiary for a given drug. Net expenditures
                                                               reflect any rebates and discounts obtained by plan sponsors and PBMS after a beneficiary receives a
                                                               drug.




                                                               Consistent with the results for all Part D drugs, from 2014 through 2016
                                                               rebates and other price concessions outpaced growth in gross and net
                                                               expenditures for the three groups of highest expenditure, highest
                                                               utilization brand-name drugs in our analysis (see table 2 for information
                                                               on these brand-name drugs). The three groups of brand-name drugs
                                                               generally had higher percent changes in rebates and other prices
                                                               concessions and in gross and net expenditures than did all Part D drugs,
                                                               which includes generics. For example, from 2014 through 2016, net
                                                               expenditures for the 200 highest expenditure brand-name drugs
                                                               increased 27 percent compared to a 13 percent increase for all Part D
                                                               drugs. Of the three groups, the 200 drugs with the highest expenditure
                                                               per utilization had the largest percentage increases in expenditures and
                                                               rebates and other price concessions. However, these drugs had relatively




                                                               Page 21                                                                GAO-19-498 Medicare Part D
                                                               low gross expenditures, rebates and other price concessions, and
                                                               utilization compared with the other two groups. 42 Increases in
                                                               expenditures for the three groups of drugs in our analysis were primarily
                                                               accounted for by increases in the price per drug rather than changes in
                                                               utilization, as indicated by the growth in expenditures exceeded growth in
                                                               their utilization.

Table 2: Percentage Change in Rebates and Other Price Concessions and Expenditures for the Highest Expenditure, Highest
Utilization Brand-Name Medicare Part D Drugs, 2014-2016

                                                                                             Median percentage growth, 2014-2016
 Drug group                                                   Gross expenditures                Rebates and other     Net expenditures                  Utilization
                                                                       (percent)                price concessions            (percent)                   (percent)
                                                                                                         (percent)
 200 Highest-expenditure                                                                38                     92                       27                         1
 200 Highest-utilization                                                                29                     61                       13                        -1
 200 Highest-expenditure per                                                            61                    302                       59                       24
 utilization
Source: GAO analysis of Centers for Medicare & Medicaid Services (CMS) data. | GAO-19-498

                                                               Notes: We analyzed CMS expenditure and rebate and other price concession information, known as
                                                               direct and indirect remuneration, for the following groups of brand-name drugs: the 200 with the
                                                               highest expenditures, the 200 with the highest utilization, and the 200 with the highest expenditures
                                                               per utilization. We identified drugs that had common ingredients, strengths, dose, and route of
                                                               administration and combined them. This resulted in a group of 444 unique brand-name drugs across
                                                               the three groups.
                                                               Gross expenditures reflect what was paid to the pharmacy by both the Part D plan sponsor,
                                                               pharmacy benefit managers on the sponsor’s behalf, and the beneficiary for a given drug. Net
                                                               expenditures reflect any rebates and other price concessions obtained by plan sponsors and PBMS
                                                               after a beneficiary receives a drug.




                                                               Net expenditures per beneficiary were similar if a Part D plan sponsor
                                                               used a PBM for rebate negotiations or if it conducted its own negotiations.
                                                               Specifically, in 2016, median net expenditures per enrollee were similar
                                                               for plan sponsors using a PBM and those that did not at $2,557 and
                                                               $2,570, respectively. Rebates and other price concessions accounted for
                                                               a median of 12 percent of gross Part D expenditures for plan sponsors
                                                               using a PBM for their negotiations and a median of 10 percent for plan




                                                               42
                                                                 The highest expenditure per utilization drugs had a median of 300 beneficiaries
                                                               receiving them in 2016, compared to a median of 53,000 and 96,000 beneficiaries for the
                                                               highest expenditure drugs and the highest utilization drugs, respectively.




                                                               Page 22                                                                GAO-19-498 Medicare Part D
                                                               sponsors that did not. 43 The majority—82 percent—of plan sponsors used
                                                               a PBM to obtain rebates and other price concessions on their behalf. 44
                                                               The plan sponsors that performed their own negotiations generally had
                                                               higher enrollment than those that used a PBM—a median of
                                                               approximately 47,000 beneficiaries, compared to approximately 13,000
                                                               beneficiaries (see table 3).

Table 3: Rebates and Other Price Concessions Received by Medicare Part D Plan Sponsors That Used a Pharmacy Benefit
Manager (PBM) for Rebate Negotiations and Those That Did Not, 2016

                                                                                                Part D plan sponsors
                                         Number of Part D                   Median Part D Rebates and other price               Median gross         Median net
                                              sponsorsa                 sponsor enrollment       concessions as                        Part D             Part D
                                                                                             percentage of gross                expenditures       expenditures
                                                                                             Part D expenditures                 per enrollee       per enrollee
                                                                                                        (median)                     (dollars)          (dollars)
 PBM-performed                                               177                            12,526                        12              3,042              2,557
 Plan-performed                                                20                           46,860                        10              3,031              2,570
Source: GAO analysis of Centers for Medicare & Medicaid Services (CMS) data. I GAO-19-498

                                                               Note: We analyzed CMS Part D expenditure and rebate and other price concessions from the
                                                               prescription drug event and direct and indirect remuneration data for 2016 to determine the amount of
                                                               rebates and other price concessions relative to gross expenditures, which reflect the amount paid to
                                                               the pharmacy by the Part D plan sponsor, PBMs on the sponsor’s behalf, and the beneficiary for a
                                                               given drug. Net expenditures reflect any rebates and discounts obtained by plan sponsors and PBMs
                                                               after a beneficiary receives a drug.
                                                               a
                                                                Excluded are 20 Part D plan sponsors for which we were unable to determine the entity performing
                                                               their rebate and price concession negotiations.


                                                               See appendix IV for additional information on expenditures and rebates
                                                               and other price concessions obtained for the 444 highest expenditure,
                                                               highest utilization brand-name Part D drugs in 2016. The appendix also
                                                               contains information on expenditures and rebates and other price
                                                               concessions obtained by the Part D plan sponsors whose representatives
                                                               we interviewed.

                                                               43
                                                                 Of the 217 Part D plan sponsors, 177 plan sponsors, accounting for 54 percent of Part D
                                                               enrollment in 2016, used a PBM to conduct their rebate and price concession
                                                               negotiations. Twenty plan sponsors, accounting for 39 percent of enrollment, performed
                                                               their own rebate and price concession negotiations. The remaining 20 sponsors, for which
                                                               we could not determine the entity conducting the negotiations, accounted for 7 percent of
                                                               the enrollment.
                                                               44
                                                                 This calculation reflects PBM use by plan sponsors. Each sponsor may have more than
                                                               one contract that offers Part D coverage. Our analysis of plan sponsor use of PBMs for
                                                               their individual Part D found that 82 percent of contracts used a PBM, either alone or in
                                                               conjunction with the plan sponsor, for rebate and price concession negotiations.




                                                               Page 23                                                               GAO-19-498 Medicare Part D
                       In 2016, the highest expenditure, highest utilization brand-name drugs
Part D Drug Prices     sold in retail pharmacies received discounts off of manufacturer list prices
Were Significantly     that were significantly higher than those sold in specialty pharmacies. 45 Of
                       the 444 highest expenditure, highest utilization brand-drugs in our
Lower Than List        analysis, 244 were sold in retail pharmacies. For this group, gross Part D
Prices for Brand-      prices—those paid to the pharmacy by the Part D plan sponsor, PBMs on
                       the sponsor’s behalf, and the beneficiary—were 17 percent lower than
Name Drugs in Retail   manufacturer list prices for these drugs. When rebates and other price
Pharmacies; Drugs      concessions were accounted for, net Part D prices were 41 percent lower
                       than manufacturer list prices. 46 In contrast, the 200 drugs sold in specialty
Sold in Specialty      pharmacies received fewer discounts off of manufacturer list prices. For
Pharmacies Received    these drugs, median gross and net prices were 15 percent and 16
                       percent, respectively, lower than manufacturer list prices (see fig. 3). 47 As
Fewer Discounts        a result, drugs sold in retail pharmacies received median discounts (41
                       percent) that were 2.5 times larger than those sold in specialty
                       pharmacies (16 percent).




                       45
                         We separated drugs sold in retail community pharmacies from those sold in specialty
                       pharmacies, as the latter dispense low-volume and high-cost drugs to patients undergoing
                       intensive therapies for illnesses.
                       46
                        Pharmacy acquisition cost, which reflects the price pharmacies paid to obtain the drug,
                       was 19 percent lower than manufacturer list prices for drugs sold in retail pharmacies.
                       47
                         Of the 200 brand-name drugs sold in specialty pharmacies, 187 were among the 200
                       Part D drugs with the highest expenditure per utilization Part D drugs in 2016. As noted
                       earlier, these drugs had lower utilization and received fewer rebates than drugs with
                       higher utilization and higher expenditures.




                       Page 24                                                        GAO-19-498 Medicare Part D
Figure 3: Medicare Part D Median Drug Prices as a Percentage of Manufacturer List
Prices for Highest Expenditure and Highest Utilization Brand-Name Drugs, 2016




Notes: Brand-name drugs sold in retail pharmacies had a gross median per unit price of $11.75, while
the gross median per unit price for brand-name drugs sold in specialty pharmacies was $315.54.
Prices are the 2016 median per unit prices for the brand-name drugs that met the following criteria:
the 200 drugs with the highest expenditures, the 200 drugs with the highest utilization, and the 200
drugs with the highest expenditures per utilization. We identified drugs that had common ingredients,
strengths, dose, and routes of administration. This resulted in a group of 444 unique brand-name
drugs across the three groups because some drugs met multiple criteria and therefore appeared in
more than one group. Of these 444 brand-name drugs, 244 were drugs sold in retail pharmacies, and
200 were drugs sold in specialty pharmacies.
Manufacturer list price is the median average wholesale price. Pharmacy acquisition cost reflects
prices reported in surveys of community retail pharmacies in the National Average Drug Acquisition
Cost data set. Gross Part D prices reflect median unit prices paid to pharmacies by Part D plan
sponsors, pharmacy benefit managers on the sponsor’s behalf, and the beneficiary and net Part D
prices account for rebates and other price concessions obtained by plan sponsors for these drugs.
a
 Pharmacy acquisition cost data are unavailable for drugs sold in specialty pharmacies, as these
pharmacies are not surveyed by CMS.




Page 25                                                               GAO-19-498 Medicare Part D
                       See appendix V for more information on prices for the highest
                       expenditure, highest utilization brand-name drugs and for information on
                       prices for selected generic drugs.


                       Our review of 52 peer-reviewed studies indicates that utilization
Utilization            management services were associated with financial savings or improved
Management Was         beneficiary health indicators. However, the effects on ensuring that
                       beneficiaries take their medication as prescribed (adherence) and access
Generally Associated   to clinically appropriate prescriptions were less clear. 48 The studies
with Financial         examined the effects of 10 different types of utilization management
                       services in three areas: (1) financial savings; (2) beneficiary health
Savings and            indicators; and (3) beneficiary medication adherence and access: 49
Improved Health
                            Financial savings. Twenty-seven of the 36 studies we reviewed that
Indicators, but Its    •
                            examined financial savings found that utilization management
Effect on Medication        services were associated with savings for the Medicare program, Part
                            D plans, or beneficiaries. For example, all eight studies that examined
Adherence and               the relationship between generic substitution and financial savings
Access Was Less             found savings. 50 Of the 10 studies that did not find financial savings,
                            five found no statistically significant impact of the utilization
Clear                       management service on savings, three found the utilization
                            management service was associated with a decrease in savings, and
                            two found both an increase and decrease in savings for different types
                            of utilization management services.
                       •    Beneficiary health indicators. Twelve of the 20 studies that
                            examined beneficiary health indicators found that utilization
                            management services were associated with improvement, such as a
                            reduction in adverse drug events. Ten of the 12 studies that found
                            improvement examined either medication therapy management
                            programs or comprehensive medication reviews. The other two
                            studies that found improvement looked at drug utilization reviews,
                            which examine a beneficiary’s prescriptions to identify safety
                            considerations, such as potential adverse interactions with other

                       48
                         The studies used a range of health indicators, such as appropriateness of medications
                       for older adults, cholesterol values, and reductions in adverse drug events.
                       49
                         Some studies examined the effect of utilization management services on more than one
                       outcome. For example, one study examined both step therapy and prior authorization.
                       50
                         Generic substitution is switching a generic drug for its bioequivalent, brand-name
                       counterpart.




                       Page 26                                                        GAO-19-498 Medicare Part D
     drugs and compliance with clinical guidelines. 51 Of the eight studies
     that found no improvement, one found that a health indicator
     worsened, and four found improvement in at least one health indicator
     and a decline in at least one other indicator.
•    Beneficiary medication adherence and access. Of the 15 studies
     that examined the effect of utilization management services on
     beneficiaries’ medication adherence or access to clinically appropriate
     drugs, 10 examined medication therapy management programs or
     comprehensive medication reviews. Seven of these 10 found
     improvement in medication adherence. 52 In contrast, the other five
     studies that examined adherence and access found negative, mixed,
     or no effects associated with prior authorization and step therapy. For
     example, two studies examined the effect of prior authorization and
     step therapy and found that these utilization management services
     resulted in increased access problems. Two other studies examined
     the relationship of prior authorization and step therapy adherence and
     found a mixed impact. The remaining study examined the relationship
     of only prior authorization with the time needed to access medications
     and found no clinically significant impact.

Stakeholders we interviewed generally agreed that utilization
management services resulted in financial savings but differed in their
views regarding the effect of utilization management services on
beneficiaries’ medication adherence and access to clinically appropriate
drugs. In interviews with representatives from PBMs, Part D plan
sponsors, and a manufacturer trade association, these stakeholders
generally agreed that utilization management services resulted in
financial savings. While representatives from most Part D plan sponsors
and PBMs told us that utilization management services have resulted in
no adverse impact on medication adherence and access to prescriptions,
representatives of the three drug manufacturers we interviewed told us
that utilization management services limit medication adherence and
access to medications by, for example, delaying therapy to needed drugs.


51
  A comprehensive medical review, which can be part of a medication therapy
management program, is a systematic process of assessing medications to identify
problems, such as a change in a drug’s effect when taken with another drug, and creating
a plan to resolve them.
52
   The remaining three studies found that the utilization management service had no
statistically significant impact on adherence.




Page 27                                                      GAO-19-498 Medicare Part D
                  See appendix VI for more information about the effects of utilization
                  management services from the peer-reviewed studies we examined and
                  the stakeholders we interviewed. See appendix VII for the articles
                  included in our literature review.


                  The Department of Health and Human Services provided technical
Agency Comments   comments on a draft copy of this report, which GAO incorporated as
                  appropriate.


                  As agreed with your offices, unless you publicly announce the contents of
                  this report earlier, we plan no further distribution until 30 days from the
                  report date. At that time, we will send copies of this report to the
                  appropriate congressional committees and the Secretary of Health and
                  Human Services. In addition, the report will be available at no charge on
                  the GAO website at http://www.gao.gov.

                  If you or your staff have any questions about this report, please contact
                  me at (202) 512-7114 or dickenj@dickenj@gao.gov.gov. Contact points
                  for our Offices of Congressional Relations and Public Affairs may be
                  found on the last page of this report. GAO staff who made major
                  contributions to this report are listed in appendix VIII.




                  John E. Dicken
                  Director, Health Care




                  Page 28                                             GAO-19-498 Medicare Part D
Appendix I: Objectives, Scope, and
              Appendix I: Objectives, Scope, and
              Methodology



Methodology

              This appendix provides details on our scope and methodology in
              addressing each of our five reporting objectives: (1) the extent to which
              Part D plan sponsors contract with pharmacy benefit managers (PBM) to
              deliver drug benefit management services to Medicare beneficiaries; (2)
              how PBMs earn revenue from the services they provide to Part D plan
              sponsors; (3) trends in rebates and other price concessions obtained by
              Part D plan sponsors and PBMs from manufacturers and others for Part
              D drugs; (4) the extent to which prices for Part D drugs are discounted off
              of manufacturer list prices; and (5) what is known about savings and other
              effects from utilization management services commonly used in Part D. In
              addition, the appendix describes the steps we took to assure the reliability
              of the data we analyzed.


Interviews    For all our objectives, we obtained the perspectives of stakeholders on
              Part D plan sponsors’ use of PBMs as well as information on sponsors’
              efforts to control Part D expenditures and drug utilization. We spoke to
              representatives from 17 small, mid-sized, and large Part D plan sponsors:
              Aetna, Anthem, Banner Health, Cambia Health, Cigna, CVS, Express
              Scripts, Kaiser, Health Care Service Corp, Health Plan of San Mateo,
              Henry Ford Health System, Humana, Missouri Highways and
              Transportation Commission, Rite Aid, United Health Care, University of
              Pittsburgh Medical Center, and WellCare. 1 We spoke with seven PBMs:
              Argus, CVS Caremark, EnvisionRx, Express Scripts, MedImpact, Prime
              Therapeutics, and OptumRx. 2 To obtain other drug industry perspectives,
              we spoke with representatives from three drug manufacturers: Eli Lilly,
              Gilead, and Amgen. We also spoke with one entity that is both a
              wholesaler and pharmacy services administrative organization:
              AmerisourceBergen. Additionally, we spoke with other industry and
              advocacy organizations, including groups representing drug

              1
               The 17 Part D plan sponsors consisted of 11 sponsors that had the 20 contracts with the
              largest enrollment in 2016 (enrollment in these contracts accounted for 82 percent of Part
              D enrollment in 2016); three plan sponsors that had contracts with enrollment at or just
              below the median 2016 contract enrollment; and three plan sponsors that had three
              contracts with enrollment at or below the bottom enrollment quartile. In this report, we
              refer to a drug plan or plans covered under each sponsor contract with the Centers for
              Medicare & Medicaid Services (CMS) as a “Part D plan sponsor contract.”
              2
               Argus is known as DST Pharmacy Solutions as of September 2017. We spoke with
              representatives from the six PBMs that provided the most drug benefit management
              services to Part D plan sponsors in 2016 as identified by our analysis of HPMS data. We
              also judgmentally selected an additional PBM that provided the eighth most drug benefit
              management services to Part D plan sponsors.




              Page 29                                                        GAO-19-498 Medicare Part D
                            Appendix I: Objectives, Scope, and
                            Methodology




                            manufacturers, Part D plan sponsors, pharmacies, and PBMs: America’s
                            Health Insurance Plans, Biotechnology Innovation Organization,
                            Community Oncology Alliance, National Association of Chain Drug
                            Stores, National Association of Specialty Pharmacies, National
                            Community Pharmacists Association, Patients for Affordable Drugs,
                            Pharmacy Benefit Management Institute, Pharmaceutical Care
                            Management Association, and Pharmaceutical Research and
                            Manufacturers of America.


The Extent to Which Part    To determine the extent to which PBMs provided services to Part D plan
D Plan Sponsors Contract    sponsors, we analyzed the Centers for Medicare & Medicaid Services’
                            (CMS) Health Plan Management System (HPMS) data that identified the
with PBMs to Deliver Drug
                            entity or entities responsible for performing each of 10 key drug benefit
Benefit Management          management services for plan sponsors’ Part D contracts in 2016, the
Services to Beneficiaries   most recent available expenditure and rebate and other price concession
                            data at the time of our analysis. 3 CMS provided HPMS data for the 624
                            Part D plan sponsor contracts that were effective in 2016. 4 The data
                            contained the entity or entities reported by each plan sponsor as
                            performing each service. Using this information, we identified for each
                            contract whether the plan sponsor performed a service itself; contracted
                            with a PBM to perform the service; or performed the service in
                            coordination with a PBM. For a given contract, we counted as being a
                            PBM any entity that was not the plan sponsor that performed one or more
                            drug benefit management services. We manually reviewed those PBMs
                            against a list of PBM members from a PBM trade organization. We used

                            3
                             HPMS is the communication portal through which Part D plan sponsors report
                            information to CMS. According to CMS, Part D plan sponsors are required to report the
                            entities responsible for performing these 10 drug benefit management services on an
                            ongoing basis: claims adjudication, rebate and other price concession negotiations,
                            pharmacy network development, enrollment processing, enrollee appeals and grievance
                            process-management, customer service, management of a pharmacy and therapeutics
                            committee, coordination with drug benefit programs, pharmacy technical assistance, and
                            drug benefit administration.
                            4
                             Plan sponsors contract with CMS to provide Part D coverage through individual contracts
                            that each offer one or more unique drug plans. Part D coverage is offered through
                            contracts providing stand-alone Part D prescription drug plans (PDPs) that supplement
                            traditional Medicare or through Medicare Advantage (Part C) plans that cover all Medicare
                            benefits, including Part D drug coverage. CMS provided us with HPMS data for all PDP
                            and Part C plan sponsor contracts. They excluded Program of All-inclusive Care for the
                            Elderly (PACE) Part D plans, as they are exempted from Part D requirements, such as
                            charging beneficiaries cost-sharing. These 624 contracts provided 4,663 unique plans and
                            were operated by 207 Part D plan sponsors in June 2016.




                            Page 30                                                      GAO-19-498 Medicare Part D
                         Appendix I: Objectives, Scope, and
                         Methodology




                         internet searches to confirm the entity was not the plan sponsor in
                         instances when it was not listed in the trade organization’s member
                         directory. In doing so, we also identified whether the plan sponsor shared
                         common ownership with the PBM responsible for providing the drug
                         benefit management service. For example, there were instances where
                         the plan sponsor and PBM were sister organizations owned by the same
                         parent company. In this situation, we counted the PBM as a separate
                         entity from the plan.

                         In addition, we analyzed PBM use by plan sponsor contract enrollment
                         size using CMS contract enrollment information from June 2016. 5
                         Additionally, we used HPMS data to examine plan sponsor contracts’
                         variation in the number of PBMs used, the types of services that PBMs
                         provided, and the use of PBMs by contract enrollment size. We also
                         identified the PBMs that provided the most services and described the
                         services they provided. Last, we interviewed Part D plan sponsor
                         representatives to understand the considerations that influenced their
                         decision about how and whether to use a PBM.


How PBMs Earn Revenue    To determine how PBMs earned revenue from services they provide to
from the Services They   Part D plan sponsors, we relied on four information sources. First, we
                         reviewed selected service agreements between PBMs and Part D plan
Provide to Part D Plan
                         sponsors. The service agreements generally contain detailed information
Sponsors                 on the services that the PBM will provide, how the plan sponsor will pay
                         the PBM for those services, and the rates that pharmacies will be paid for
                         Part D drugs. We asked CMS for a list of all service agreements it
                         approved between January 2016 and May 2018 that were in effect as of
                         June 2018. CMS provided us with a list of 119 service agreements. Using
                         June 2018 Part D publicly available enrollment data from CMS, we
                         obtained from CMS the 20 service agreements for Part D plans sponsors
                         with the largest enrollment in June 2018. While most of the service
                         agreements included sufficient information to determine how the PBMs
                         were paid, some did not, and, where appropriate, we noted these
                         instances in our findings.

                         Second, we examined PBM revenue reported to CMS by Part D plan
                         sponsors in their rebates and other price concession data—also referred

                         5
                          We used June 2016 data to indicate enrollment, as this month has relatively stable
                         enrollment as it does not fall within the annual open enrollment period where beneficiaries
                         may change their Part D coverage.




                         Page 31                                                        GAO-19-498 Medicare Part D
                           Appendix I: Objectives, Scope, and
                           Methodology




                           to as direct and indirect remuneration (DIR)—for 2014, 2015, and 2016.
                           These rebate and other price concession submissions contain information
                           on the various sources of revenue and expenses incurred by PBMs and
                           plan sponsors. 6

                           Third, we reviewed applicable CMS regulations and guidance on the
                           reporting of PBM and Part D plan sponsor revenue and expenses.

                           Fourth, we interviewed PBM representatives about the extent to which
                           PBMs retained rebates or passed them through to plan sponsors and, in
                           some cases, the reasons for this decision. We also asked certain PBM
                           representatives whether their revenue sources for Part D, specifically
                           rebate retention and spread pricing, differed from PBMs’ and plan
                           sponsors’ commercial business and, if so, the reasons for any
                           differences.


Rebates and Other Price    To examine rebates and other price concessions obtained by Part D plan
Concessions Obtained by    sponsors and PBMs from manufacturers and others for Part D drugs,
                           relative to overall Part D expenditures, we analyzed plan sponsors’ gross
Part D Plan Sponsors and
                           and net expenditures for Part D drugs for 2014 through 2016, the most
PBMs from Manufacturers    recent data available at the time of our analysis. Gross expenditures
and Others for Part D      reflect what was paid to the pharmacy by the plan sponsor, PBMs on the
Drugs                      sponsor’s behalf, and the beneficiary for a given drug. Net expenditures
                           reflect any rebates and other price concessions obtained by Part D plan
                           sponsors and PBMs after a beneficiary receives a drug. To calculate
                           gross expenditures, we used Medicare prescription drug event (PDE)
                           data to calculate gross brand-name and generic drug expenditure and
                           utilization information for all Part D plan sponsors’ contracts. 7 We used

                           6
                            We reviewed Part D plan-sponsor-filed DIR data, which contains information on rebates
                           and other price concessions, along with additional breakdowns of other revenue sources,
                           including monies retained by PBMs for non-rebate services they provided to
                           manufacturers and spread pricing—the difference between what the PBM paid the
                           pharmacy and charged the Part D plan sponsor for a drug.
                           7
                            Part D plan sponsors—both PDPs and Medicare Advantage drug plans—must submit a
                           PDE record to CMS each time a beneficiary obtains a prescription drug. The PDE record
                           contains information on the beneficiary receiving the drug, the price paid by the plan
                           sponsor to the pharmacy, and applicable beneficiary cost-sharing. We excluded PDE
                           claims billed under PACE contracts because they are exempted from certain Part D
                           requirements, such as charging beneficiaries cost-sharing. We also excluded
                           compounded drugs, which are tailor-made by a pharmacy for a beneficiary, and over-the-
                           counter drugs as they are generally not covered by Medicare Part D.




                           Page 32                                                      GAO-19-498 Medicare Part D
Appendix I: Objectives, Scope, and
Methodology




Red Book, a compendium published by Truven Health Analytics, to
determine whether drugs were brand-name or generic. 8 We then
identified individual brand-name and generic drugs by grouping
expenditure claims with the same active ingredient, strength, dosage
form, and route of administration (known as ISDR). 9 We calculated brand-
name and generic drug expenditures based on a drug’s ingredient cost,
dispensing fees, sales tax, and applicable vaccine administration fees.
We used PDE data to calculate gross expenditures for all Part D plan
sponsors at both the contract and plan sponsor level. 10 We used DIR data
to determine the amount of rebates and other price concessions and
subtracted this amount from this data to calculate net expenditures. 11 We
also obtained plan sponsor enrollment data using publicly available CMS
data for June 2016, which allowed us to calculate gross per beneficiary
expenditures. 12

We also examined differences in the amount of rebate and other price
concessions obtained relative to expenditures for Part D plan sponsors
that used a PBM relative to those that did not. We determined PBM
involvement in rebate and other price concession negotiations for
individual plan sponsors using 2016 HPMS data. We specifically looked
at each entity listed in HMPS as negotiating rebates and other price
concessions with drug manufacturers and others. We were able to
determine whether a PBM or plan sponsor performed this service for 197
plans sponsors. However, there were 20 Part D plan sponsors where a

8
 Generic drugs may have more than one manufacturer, and we grouped these drugs
together regardless of whether they were produced by more than one manufacturer.
9
 The dosage form is the physical form in which a drug is produced and dispensed, such
as a tablet or capsule; route of administration is the way of administering a drug to a
patient, such as taking a drug orally. We used the Red Book to determine drugs’
therapeutic class, dose, and route of administration.
10
  Part D plan sponsors may have one or more unique contracts that each offer one or
more unique drug plans. We collapsed all sponsors’ respective contract expenditure
information, which resulted in a list of 217 unique plan sponsors in 2016.
11
  Rebates are a form of price concession paid by a drug manufacturer to the plan sponsor
or the PBM working on the plan’s behalf generally after a drug was purchased by a
beneficiary. Plan sponsors may receive other price concessions that lower the price of a
drug. For example, plan sponsors may receive fees from pharmacies based on their
performance.
12
  We used June 2016 data to indicate enrollment because this month has relatively stable
enrollment as it does not fall within an annual open enrollment period where beneficiaries
may change their Part D coverage.




Page 33                                                       GAO-19-498 Medicare Part D
Appendix I: Objectives, Scope, and
Methodology




PBM or plan was not solely listed as performing the rebate and other
price concession service. In these instances, we could not identify which
entity negotiated rebates and other price concessions and therefore
excluded them from this analysis. 13

To obtain more information on drugs that have the greatest fiscal impact
on the Part D program and beneficiaries, we calculated gross and net
expenditures for the brand-name and generic drugs with the highest
expenditures, highest utilization, and highest expenditure per utilization in
2016. For both brand-name and generic drugs, we identified the following:
the 200 brand-name and 200 generic drugs with the highest expenditures
in 2016; the 200 brand-name and generic drugs with the highest
utilization in 2016 (based on number of 30-day prescriptions); and the 200
brand-name and generic drugs with the highest expenditures per
utilization (i.e., highest expenditure per number of 30-day prescriptions).
As a result of overlap in the groups of drugs, these criteria yielded two
groups: the 444 highest expenditure, highest utilization brand-name drugs
and the 476 unique highest expenditure, highest utilization generic drugs.
These 920 drugs accounted for 81 percent of total Part D expenditures in
2016. 14 We used drug-level rebate and other price concessions data to
calculate net drug prices for these drugs by subtracting rebate and other
price concessions for each drug from gross expenditures. 15




13
  We excluded instances in which the plan sponsor and PBM were listed as performing
this service together because we wanted to isolate any differences in expenditures and
the amount of rebates and other price concessions based on whether the plan sponsor or
PBM performed the service. This differs from our analysis of Part D plan sponsors’ use of
PBMs in the Part D program, where we examined the extent to which PBMs were involved
in providing these services, regardless of whether the PBM performed alone in
conjunction with the plan sponsor.
14
  Of the 920 highest expenditure, highest utilization brand-name and generic drugs, the
444 brand-name drugs accounted for 65 percent of total drug expenditures and the 476
generic drugs accounted for 16 percent of expenditures.
15
  In addition to plan sponsor-level DIR data, Part D plan sponsors submit drug-level DIR
data to CMS. This contains information on the amount of rebates and other price
concessions, but does not include the additional breakdowns included in the plan sponsor-
level data (e.g., information on retained rebates).




Page 34                                                       GAO-19-498 Medicare Part D
                           Appendix I: Objectives, Scope, and
                           Methodology




The Extent to Which        To determine the extent to which Part D drug prices are discounted off of
Prices for Part D Drugs    manufacturer list prices, we compared the median gross and net prices
                           for the 444 brand-name and 476 generic highest expenditure, highest
Are Discounted Off of
                           utilization drugs to (1) list prices established by manufacturers, and (2)
Manufacturer List Prices   the cost to pharmacies of acquiring these drugs. For list prices, we used
                           2016 average wholesale price (AWP) data from Truven Health Analytics’
                           Red Book. 16 AWP is a common benchmark drug price used in the
                           negotiation of payment rates between Part D plan sponsors and
                           pharmacies. 17 Because AWP is updated on an ongoing basis, we
                           calculated a day-weighted per unit price that takes into account the
                           number of days that the reported price was in effect in 2016. We then
                           determined the median AWP price for each drug product based on the
                           ISDR. 18 We refer to the median price as the manufacturer list price.

                           For pharmacy acquisition costs, which reflect the price pharmacies paid
                           to obtain the drug, we used retail community pharmacy acquisition cost
                           data from National Average Drug Acquisition Cost (NADAC) data.
                           NADAC does not contain data from non-retail pharmacies, such as mail-
                           order or specialty pharmacies. For our groups of 444 brand-name and
                           476 generic drugs, we separated drugs sold in retail community
                           pharmacies from those sold in specialty pharmacies. If a drug did not
                           have pharmacy acquisition cost data from NADAC, we considered that
                           drug to be sold in specialty pharmacies and, thus, a specialty drug. 19

                           We used 2016 PDE data to determine the gross per unit Part D price for a
                           drug by dividing the gross expenditures for the drug by the total quantity




                           16
                             We analyzed 2016 AWP data, as this was the most recent data for which both PDE and
                           DIR data were available at the time of our analysis.
                           17
                             AWP is the manufacturer’s suggested price for a drug sold by a wholesaler to a
                           pharmacy and does not reflect the actual price charged by a wholesaler.
                           18
                             A drug with the same proprietary name may have more than one ISDR. For example,
                           Amitiza, which is lubiprostone, has separate ISDR for 8 microgram and 24 microgram
                           capsules.
                           19
                             We determined whether a drug was sold in retail or specialty pharmacies based on
                           NADAC data. We found that 97 percent of the 444 highest expenditure, highest utilization
                           brand-name drugs that lacked NADAC were listed as specialty drugs on a Part D plan
                           sponsor’s formulary. Given this, we refer to drugs that do not have a NADAC price as
                           those sold in specialty pharmacies.




                           Page 35                                                      GAO-19-498 Medicare Part D
                            Appendix I: Objectives, Scope, and
                            Methodology




                            dispensed of it. 20 For example, a drug that had 1,000 units prescribed to
                            Medicare beneficiaries and $5,000 in gross expenditures would have a
                            gross per unit price of $5. We determined net per unit Part D prices for
                            the drugs in our two study groups by dividing the amount of rebates and
                            other price concessions for each drug by the quantity dispensed of it and
                            then subtracting the amount of rebates and other price concessions per
                            quantity from the gross Part D price for each drug.

                            For each drug, we then determined the median pharmacy acquisition cost
                            (if available), median gross Part D price, and median net Part D price as a
                            proportion of median manufacturer list price by dividing each price by the
                            median manufacturer list price. We then reported the median value for
                            these pricing points for the highest expenditure, highest utilization drugs
                            in our analysis.


Analysis of Literature on   To determine what is known about the impact of utilization management
Effect of Utilization       services that PBMs commonly provide to Part D plan sponsors, or that
                            plan sponsors may perform themselves, we conducted a literature search
Management Services
                            for studies that examined the effect of utilization management services in
                            Part D (regardless of whether they were provided by a PBM or another
                            entity) on the following outcomes: (1) financial costs or savings, (2)
                            beneficiaries’ health indicators, and (3) beneficiaries’ access to clinically
                            appropriate medications or taking their medications as prescribed
                            (adherence). 21 The literature search was performed from April 2018 to
                            July 2018 using keyword searches in bibliographic databases, including
                            ProQuest, EBSCO, and Scopus. We limited our search to studies
                            published beginning in 2006—the year the Part D program began.

                            For our searches, we developed a list of search terms for our literature
                            review by reviewing relevant background documentation and several
                            database searches. The search terms included: “utilization management,”
                            “prior authorization,” “quantity limits,” “step therapy,” “generic
                            substitution,” “drug utilization review,” “quantity edit,” “medication therapy
                            management,” and “comprehensive medication review,” combined with

                            20
                              Quantity dispensed is measured in units such as milliliters or milligrams. For a given
                            drug, we summed the total number units given to Medicare beneficiaries and divided it by
                            the gross expenditures for the drug.
                            21
                              Utilization management services may be included under various combinations of the 10
                            prescription drug benefit management services, instead of being a distinct service that
                            Part D plan sponsors report to CMS in its HPMS database.




                            Page 36                                                      GAO-19-498 Medicare Part D
                   Appendix I: Objectives, Scope, and
                   Methodology




                   “access,” “adherence,” “health benefit,” “clinical outcome,” “generic use,”
                   “cost effectiveness,” “savings,” “costs,” and “Medicare.”

                   The literature search generated 700 studies. We reviewed this list by
                   examining the abstracts for those studies that addressed the effects of
                   utilization management services in Part D and were published in peer-
                   reviewed journals. We identified 48 studies that met our criteria then
                   added four more that met the criteria from several literature reviews we
                   examined, resulting in a final group of 52 peer-reviewed studies that we
                   analyzed. We analyzed these studies to group them by type of utilization
                   management service evaluated and type of outcome measured. We
                   documented any methodological limitations of these studies but did not
                   exclude any of them on this basis. See the bibliography in Appendix VII
                   for a list of the 52 studies in our review.

                   We also interviewed PBMs, plan sponsors, and drug manufacturers to
                   obtain their views regarding the impact of utilization management
                   services in Part D plans and asked them to recommend additional studies
                   on utilization management services. We did not assess the methodology
                   or data reliability of the studies provided to us by these drug supply chain
                   stakeholders; none of them met our criterion of being published in peer-
                   reviewed journals. We used these studies to better understand
                   stakeholder perspectives.


Data Reliability   To ensure the data used to produce this report were sufficiently reliable,
                   we took several steps. We performed data reliability checks on the HPMS
                   data by reviewing the data for missing values and errors, checking the
                   information against other publicly available sources, and interviewing
                   knowledgeable agency officials. We performed data reliability checks on
                   the PDE and DIR data by reviewing relevant documentation, checking the
                   data for outliers and errors, and interviewing knowledgeable agency
                   officials. We performed data reliability checks of the AWP and NADAC
                   data sets by testing the data for missing data and outliers and reviewing
                   relevant documentation. After taking these steps, we determined the data
                   were sufficiently reliable for the purposes of our reporting objectives.

                   We conducted this performance audit from May 2017 to July 2019 in
                   accordance with generally accepted government auditing standards.
                   Those standards require that we plan and perform the audit to obtain
                   sufficient, appropriate evidence to provide a reasonable basis for our
                   findings and conclusions based on our audit objectives. We believe that



                   Page 37                                              GAO-19-498 Medicare Part D
Appendix I: Objectives, Scope, and
Methodology




the evidence obtained provides a reasonable basis for our findings based
on our audit objectives.




Page 38                                           GAO-19-498 Medicare Part D
Appendix II: Medicare Part D Plan Sponsors’
                            Appendix II: Medicare Part D Plan Sponsors’
                            Use of Pharmacy Benefit Managers (PBM)



Use of Pharmacy Benefit Managers (PBM)

                            This appendix provides additional detail on the use of PBMs by Part D
                            plan sponsors to provide prescription drug benefit management services
                            to Medicare beneficiaries.


Part D Plan Sponsors’ Use   We examined Centers for Medicare & Medicaid Services’ (CMS) data to
of PBMs                     identify the 10 key drug benefit management services provided by PBMs
                            under 624 Part D plan sponsor contracts in 2016, the most recent
                            available expenditure and rebate and other price concession data at the
                            time of our analysis, and found the following variation in plan sponsor use
                            of PBMs: 1

                            •   Services provided by PBMs. Part D plan sponsors’ contracts varied
                                by the services provided by PBMs in 2016. Plan sponsors’ use of a
                                PBM for drug benefit services—either alone or with the plan
                                sponsor—for their 624 contracts varied from 30 percent for enrollee
                                appeals and grievance process-management to 99 percent for claims
                                adjudication. For seven of the 10 drug benefit management services,
                                PBMs—either alone or in conjunction with the plan sponsor—provided
                                services to more than half the sponsor contracts (see fig. 4).




                            1
                             In this report, we refer to a drug plan or plans covered under each sponsor contract with
                            CMS as a “Part D plan sponsor contract.”




                            Page 39                                                        GAO-19-498 Medicare Part D
                                       Appendix II: Medicare Part D Plan Sponsors’
                                       Use of Pharmacy Benefit Managers (PBM)




Figure 4: Medicare Part D Plan Sponsor Contract Use of Pharmacy Benefit Managers (PBM) for 10 Drug Benefit Management
Services, by Percent of Contracts, 2016




                                       Note: We analyzed CMS Health Plan Management System data for 2016 that identified the entity or
                                       entities responsible for performing each of 10 key drug benefit management services for the 624 Part
                                       D plan sponsor contracts in effect in 2016. We excluded contracts that participated in the Medicare
                                       Program of All-Inclusive Care because they are exempted from Part D requirements such as charging
                                       beneficiaries cost-sharing.


                                       •    Number of PBMs used. Part D plan sponsor contracts varied in the
                                            number of PBMs used to provide one or more of the 10 drug benefit
                                            management services. For example, 54 percent of plan sponsors’



                                       Page 40                                                              GAO-19-498 Medicare Part D
Appendix II: Medicare Part D Plan Sponsors’
Use of Pharmacy Benefit Managers (PBM)




     contracts used a single PBM, while 11 percent used four or more
     PBMs (see fig. 5).


Figure 5: Number of Pharmacy Benefit Managers (PBM) Used for 10 Drug Benefit
Management Services by Percent of All Medicare Part D Plan Sponsor Contracts,
2016




Note: We analyzed CMS Health Plan Management System data for 2016 that identified the entity or
entities responsible for performing each of 10 key drug benefit management services for the 624 Part
D plan sponsor contracts in effect in 2016. We excluded contracts that participated in the Medicare
Program of All-Inclusive Care because they are exempted from certain Part D requirements, such as
charging beneficiaries cost-sharing.


•    Use of PBMs by enrollment. Smaller Part D contracts—those with
     contract enrollment below the median enrollment of all Part D
     contracts—used a PBM more often than larger contracts—those with
     enrollment at or above the median. For instance, 87 percent of
     smaller Part D plan sponsor contracts used a PBM alone or with the
     plan sponsor for rebate and price concession negotiations, compared
     to 77 percent of larger contracts. Similarly, 54 percent of smaller Part
     D contracts used a PBM alone or with a plan for a pharmacy and
     therapeutics committee, compared to 35 percent of larger contracts.
•    Use of financially related PBMs. Part D plan sponsors’ contracts
     varied by their use of PBMs with which they were related by common




Page 41                                                              GAO-19-498 Medicare Part D
                             Appendix II: Medicare Part D Plan Sponsors’
                             Use of Pharmacy Benefit Managers (PBM)




                                 ownership—either as a subsidiary or a sister company. 2 In 2016, plan
                                 sponsors used a PBM with which they were related by common
                                 ownership for 17 percent of the 624 Part D plan sponsors contracts.
                                 Larger contracts—those with enrollment at or above the median—
                                 were more likely to use a PBM related by common ownership than
                                 smaller contracts. Larger contracts used a financially related PBM for
                                 24 percent of drug benefit management services, compared to 10
                                 percent of drug benefit management services provided to smaller
                                 contracts.

Factors Influencing Plans’   The Part D plan sponsor representatives with whom we spoke noted
Decisions to Use a PBM       several considerations that influenced their decision about how and
                             whether to use a PBM. One plan sponsor noted that small plans may lack
                             the resources to conduct their own rebate negotiations and, therefore,
                             may use a PBM instead. Three other plan sponsors noted they switched
                             from conducting their own rebate negotiation with manufacturers to using
                             a PBM. Two plan sponsors said this switch was due to PBMs’ ability to
                             obtain larger rebates than the plan sponsor could, and the third
                             determined a PBM would help it achieve the best value and quality, while
                             meeting Part D’s regulatory requirements.

                             In contrast, representatives of three other Part D plan sponsors noted
                             advantages of performing drug benefit management services themselves.
                             For example, one plan sponsor noted that it performs almost all drug
                             benefit management services internally, as it believes doing so improves
                             quality through better communication and care coordination with
                             pharmacies. Another plan sponsor noted the decision not to contract out
                             certain services to a PBM may be influenced by a desire for more
                             customization over formulary management and greater control over prior
                             authorization. Representatives of one plan sponsor noted that their plan
                             does not use a PBM because they believe they are more effective in
                             developing formularies with better utilization management and greater
                             use of generic drugs than are PBMs. 3




                             2
                              For example, UnitedHealth Group owns UnitedHealthcare—which operates Part D
                             plans—and OptumRx—a PBM.
                             3
                              Greater use of generic drugs is associated with financial savings, as generics are
                             generally less expensive than brand-name drugs.




                             Page 42                                                        GAO-19-498 Medicare Part D
                                                               Appendix II: Medicare Part D Plan Sponsors’
                                                               Use of Pharmacy Benefit Managers (PBM)




Variation in the Number of                                     Our analysis of CMS data for the 624 Part D plan sponsor contracts found
Services Provided by                                           that the five PBMs that provided the largest number of services to Part D
                                                               plan sponsors’ contracts in 2016 also generally provided a full range of
PBMs to Part D Plan
                                                               PBM services to them. Four of the top five PBMs provided all 10 drug
Sponsors                                                       benefit management services to plan sponsors’ contracts while the fifth
                                                               PBM conducted claims adjudication but used an intermediary to conduct
                                                               rebate negotiations. 4 (See table 4). Furthermore, the top five PBMs
                                                               provided a high proportion of the services that Part D plan sponsors most
                                                               commonly used a PBM to provide. For example, CVS Caremark provided
                                                               claims adjudication to 144 (23 percent) of Part D plan sponsor contracts,
                                                               and OptumRx provided this service to 138 (22 percent).

Table 4: Number of Medicare Part D Plan Contracts That Used Top Five Pharmacy Benefit Managers (PBM) for Drug Benefit
Management Services, 2016

                                                                          Number of contracts that used each PBM for the service (n=624)
Drug benefit management service                                CVS Caremark                 OptumRx   Express Scripts          MedImpact                  Argusa
Claims adjudication                                                             144             138                  93                   73                    74
Drug benefit administration                                                     145             134                  92                   73                    73
Coordination with drug benefit                                                  137             120                  89                   67                    61
programs
Customer service                                                                 91              87                  30                   49                     9
Pharmacy network development                                                    142             130                  93                   73                    29
Enrollment processing                                                            14               3                   3                    8                     0
Management of a pharmacy and                                                     79              18                  21                   50                     0
therapeutics committee
Rebate and price concession                                                     119             120                  86                   59                     3
negotiation
Enrollee appeals and grievance                                                   44              11                  17                   25                     0
management
Pharmacy technical assistance                                                   124             132                  91                   69                    66
Source: GAO analysis of Centers for Medicare & Medicaid Services (CMS) data. | GAO-19-498

                                                               Note: We analyzed CMS Health Plan Management System data for 2016 that identified the entity or
                                                               entities responsible for performing each of 10 key drug benefit management services for the 624 Part
                                                               D plan sponsor contracts in effect in 2016. We excluded contracts that participated in the Medicare
                                                               Program of All-Inclusive Care because they are exempted from Part D requirements, such as
                                                               charging beneficiaries cost-sharing. In some instances, a Part D plan sponsor used more than one
                                                               PBM to provide a drug benefit management service. In these instances, multiple PBMs were counted
                                                               as providing the same service in our summary counts.
                                                               a
                                                                Argus is known as DST Pharmacy Solutions as of September 2017.


                                                               4
                                                                This PBM did not provide enrollment processing or appeals and grievance process-
                                                               management.




                                                               Page 43                                                              GAO-19-498 Medicare Part D
Appendix II: Medicare Part D Plan Sponsors’
Use of Pharmacy Benefit Managers (PBM)




In contrast, we found that Part D plan sponsors used a large number of
PBMs to provide a limited range of drug benefit management services.
For example, 48 percent of PBMs provided only one type of drug benefit
management service to plan sponsors’ contracts, and 22 percent of
PBMs provided only one service to only one plan sponsor contract. For
instance, there were 10 unique entities counted as PBMs in our analysis
that provided only customer service support to one plan sponsor contract.
One PBM representative noted in an interview that it is relatively common
for plan sponsors and PBMs to contract with other vendors to provide
additional assistance with drug benefit management services. One plan
sponsor told us, for example, that its PBM uses a vendor to manage
customer service calls.




Page 44                                           GAO-19-498 Medicare Part D
Appendix III: Information on Pharmacy
                           Appendix III: Information on Pharmacy Benefit
                           Manager (PBM) Revenue Earned from
                           Manufacturers and from Spread Pricing


Benefit Manager (PBM) Revenue Earned
from Manufacturers and from Spread Pricing
                           This appendix provides additional detail on (1) non-rebate revenue that
                           PBMs may earn for services provided to manufacturers and Medicare
                           Part D plan sponsors, and (2) PBM perspectives on Centers for Medicare
                           & Medicaid Services (CMS) policies relating to spread pricing in Part D.


Non-rebate Revenue That    PBMs and Part D plan sponsors may earn non-rebate revenue from
PBMs May Earn for          manufacturers for providing certain services. Even though this money is
                           reported to CMS as part of the rebate and other price concession
Services Provided to
                           submission, not all of it is considered rebates or other price concessions,
Manufacturers and Part D   which will lower plan liability in determining bids and thereby lower
Plan Sponsors              premiums. Of the $516.5 million in non-rebate revenue paid by
                           manufacturers in 2016, $440 million, or about 85 percent, represented the
                           amount paid for the services that exceeded the fair market value of the
                           service and is considered rebates and other price concessions. These
                           may be used to reduce the drug costs incurred by the plan sponsor.
                           Therefore, this revenue factors into bid determinations and may be used
                           to reduce premiums.

                           The remaining $78.6 million in payments from manufacturers were
                           considered “bona fide service fees”—fees paid by manufacturers to Part
                           D plan sponsors and PBMs for services that the manufacturer would
                           otherwise perform, or contract for, and that represented the fair market
                           value of those services. Such fees do not reduce the plan sponsor’s drug
                           costs and, therefore, could not factor into reducing premiums. 1 The
                           determination of a bona fide service fee as reported to CMS is made by
                           the drug manufacturer and the Part D plan sponsor and is not routinely
                           evaluated by CMS, agency officials told us. However, CMS requires that
                           the PBM and manufacturer have information documenting the fair market
                           value of the service.


Stakeholder Perspectives   CMS requires Part D plan sponsors to report revenue earned from
on PBM Revenue Earned      rebates retained by the PBM. This revenue increases the plan’s liability,
                           which increases the amount of plan bids and, therefore, result in higher
from Spread Pricing,
                           premiums. In contrast, rebate revenue passed through by PBMs to Part D
Rebate Retention, and      plan sponsors lowers the plan’s liability, reduces plans bids, and,
Differences from the       therefore, lowers beneficiary premiums.
Commercial Sector
                           1
                           See 74 Fed. Reg. 1494, 1512 (Jan. 12, 2009) (preamble, II.B.4.a.).




                           Page 45                                                     GAO-19-498 Medicare Part D
Appendix III: Information on Pharmacy Benefit
Manager (PBM) Revenue Earned from
Manufacturers and from Spread Pricing




Some PBMs earn more revenue from spread pricing in their commercial
business than in Part D, officials from three PBMs told us. Officials from
two of these PBMs noted that CMS requirements create a disincentive to
engage in spread pricing that is not present in the commercial sector.
Beginning in 2010, CMS required that plan sponsors base the amount of
beneficiary cost-sharing on the amount received by the pharmacy for a
drug—known as the “pass-through price.” 2 CMS also required that an
estimate of rebates or other price concessions be included in the
administrative costs submitted by the plan sponsor for bid
determinations. 3 Part D plan sponsors can still agree to pay the PBM
based on the higher price of the drug without accounting for rebates,
known as the lock-in price. However, the difference between that amount
and the pass-through price would increase the bid determination and
ultimately increase the premiums that plans charge beneficiaries. 4
Because there are no similar requirements pertaining to the commercial
prescription drug benefit market, spread pricing is more common there,
CMS officials told us.




2
 See 74 Fed. Reg. 1505, 1543 (Jan. 12, 2009) (preamble, II.B.3.b.) (codified at 42 C.F.R.
§ 423.100) (2018) (amending definition of “negotiated prices” upon which cost sharing is
based to include reductions due to price concessions).
3
 See 74 Fed. Reg. 1512, 1513, 1544 (Jan. 12, 2009) (preamble II.B.4.a., b.)(codified at 42
C.F.R. § 423.308 (2018)) (amending definitions of “actually paid” and “administrative
costs” to provide that rebates or price concessions are administrative costs, which will be
included in the determination of bid amounts).
4
See 42 C.F.R. §§ 423.265 (c) (2018) (costs included in bids).




Page 46                                                         GAO-19-498 Medicare Part D
Appendix IV: Expenditures and Rebate and
                             Appendix IV: Expenditures and Rebate and
                             Other Price Concession Information for
                             Medicare Part D Drugs


Other Price Concession Information for
Medicare Part D Drugs
                             This appendix provides information on (1) pharmacy-related price
                             concessions for all Medicare Part D drugs and (2) expenditure and rebate
                             and other price concession information for the 444 highest expenditure,
                             highest utilization brand-name Part D drugs in 2016. The appendix also
                             contains additional information on expenditures and rebates and other
                             price concessions obtained by the 16 Part D plan sponsors whose
                             representatives we interviewed.


Pharmacy-Related Price       The amount of pharmacy-related price concessions obtained by Part D
Concessions for All Part D   plan sponsors, or pharmacy benefit managers (PBM) on plan sponsors’
                             behalf, increased 295 percent from 2014 through 2016, from $538 million
Drugs
                             to $2.1 billion (see fig. 6). These monies account for any adjustments to
                             the price of the drug paid to the pharmacy after the point sale, such as a
                             pharmacy returning money that was overpaid by the plan sponsor or vice
                             versa. It can also include monies paid based on pharmacies’ performance
                             in meeting agreed-upon performance metrics—for example, fees a
                             pharmacy pay plan sponsors, or bonuses pharmacies receive from plan
                             sponsors, based on their performance. 1 In 2016, Part D plan sponsors
                             received $2.3 billion from pharmacies and paid out $211 million, for a net
                             of $2.1 billion in pharmacy-related price concessions.




                             1
                              An example of a performance metric is a specified percent of prescriptions dispensed for
                             a generic drug (instead of a brand-name drug).




                             Page 47                                                       GAO-19-498 Medicare Part D
Appendix IV: Expenditures and Rebate and
Other Price Concession Information for
Medicare Part D Drugs




Figure 6: Pharmacy-related Price Concessions, 2014-2016




Note: We analyzed 2014 through 2016 price concession data contained in the direct and indirect
remuneration data submitted by Medicare Part D plan sponsors to CMS. Specifically, we analyzed
monies paid to, and received from, pharmacies that affect the price of a drug paid by a Part D plan
sponsor and pharmacy benefit manager on its behalf following the purchase from a pharmacy.




Five of the seven PBMs and seven of the 12 Part D plan sponsors whose
representatives we interviewed said they have performance-based
arrangements with pharmacies. One plan sponsor noted that its
performance agreement involves paying bonuses to pharmacies that
exceed performance measures, while charging fees to pharmacies that
did not meet the measures. The sponsor said this is part of an attempt to
move from paying for volume to paying for value. Another plan sponsor
told us there has been an improvement in pharmacy performance as a
result of the program.

Representatives from pharmacy industry groups said these pharmacy-
related fees have put increasing pressure on pharmacies. For example,
one group noted there is no standardization across measures with each
plan sponsor using its own measures, and it is difficult for pharmacies to
tie a fee to a specific pharmacy location or claim. Another group noted
that fees may be imposed on pharmacies for performance measures not
directly applicable to the pharmacy. For example, the group said specialty



Page 48                                                                GAO-19-498 Medicare Part D
                         Appendix IV: Expenditures and Rebate and
                         Other Price Concession Information for
                         Medicare Part D Drugs




                         pharmacies have been assessed fees for beneficiary lack of adherence to
                         maintenance medications, such as blood pressure medications, that
                         these pharmacies do not commonly provide.


Expenditure and Rebate   PBMs and Part D plan sponsors obtained rebates and other price
and Other Price          concessions for 441 (99 percent) of the 444 highest-expenditure, highest-
                         utilization brand-name drugs in 2016. The amount of rebates and other
Concession Information
                         price concessions for each drug ranged from $1,300 to $1.8 billion in
for 444 Highest          2016, with a median of $3.3 million. 2 Rebates accounted for $24.5 billion
Expenditure, Highest     of the $26 billion in rebates and other price concessions (94 percent)
Utilization Brand-Name   obtained by plan sponsors and PBMs for these 444 drugs. As a
Drugs                    proportion of gross Part D expenditures—the amount paid by plan
                         sponsors, or the PBM on the sponsors’ behalf, and by beneficiaries—for
                         the 444 drugs ranged from -0.5 percent to 70.5 percent. 3 (See fig. 7.)




                         2
                          Three of the 444 drugs had “negative rebates and other price concessions” ranging from
                         -$10,000 to -$37,000 in 2016. According to CMS, negative rebate and other price
                         concession amounts may occur when, for example, risk-sharing arrangements between a
                         plan sponsor and a physician network resulted in more in bonuses paid to the network for
                         reducing drug costs than was received from manufacturers in rebates.
                         3
                          Gross expenditures do not account for rebates and other price concessions. Generic
                         drugs received few rebates and other price concessions compared to brand-name drugs.
                         In 2016, all 476 generic drugs in our analysis received rebates and other price
                         concessions, ranging from $48 to $8.9 million for a drug. The amount of rebates and other
                         price concessions obtained for these drugs accounted for 2.3 percent ($547 million) of the
                         $23.9 billion spent on these drugs in 2016. Of the 476 generic drugs, 379 received
                         rebates, which ranged from $1 to $2 million for a drug. Across the 920 highest
                         expenditure, highest utilization brand-name and generic drugs in our analysis, generic
                         drugs received 0.13 percent of the $24.6 billion in rebates and 26 percent ($515 million) of
                         the $2 billion in price concessions.




                         Page 49                                                         GAO-19-498 Medicare Part D
Appendix IV: Expenditures and Rebate and
Other Price Concession Information for
Medicare Part D Drugs




Figure 7: Rebates and Other Price Concessions Received for the Highest-
Expenditure, Highest-Utilization Medicare Part D Brand-Name Drugs as a
Proportion of Their Gross Expenditures, 2016




Note: We analyzed CMS expenditure and rebate and other price concession data from the
prescription drug event and direct and indirect remuneration DIR data sets for three groups of brand-
name drugs: the 200 with the highest expenditures, the 200 with the highest utilization, and the 200
with the highest expenditures per utilization (i.e., highest expenditure per number of 30-day
prescriptions). We identified drugs that had common ingredients, strengths, dose, and route of
administration and combined them. This resulted in a group of 444 unique brand-name drugs across
the three groups.




Page 50                                                               GAO-19-498 Medicare Part D
                                         Appendix IV: Expenditures and Rebate and
                                         Other Price Concession Information for
                                         Medicare Part D Drugs




Expenditures and Rebates                 Expenditures and rebates and other price concessions varied by
and Other Price                          therapeutic class for the 444 highest expenditure, highest utilization drugs
                                         in 2016. Among those with 10 or more drugs in their class, gross
Concessions for the 444
                                         expenditures ranged from $2.9 billion to $21.2 billion, and rebates and
Highest Expenditure,                     other price concessions ranged from $170 million to $8.7 billion (see table
Highest Utilization Part D               5). Four classes—endocrine metabolic agents, anti-infective agents,
Brand-Name Drugs, by                     respiratory agents, and central nervous system agents—accounted for 54
Therapeutic Drug Class,                  percent of the gross Part D expenditures, and 62 percent of rebates and
                                         other price concessions for the 444 highest expenditure, highest
2016                                     utilization drugs. 4 When accounting for rebates and other price
                                         concessions, these drugs accounted for 51 percent of net Part D
                                         expenditures.

Table 5: Expenditures and Rebates and Other Price Concessions for the 444 Highest Expenditure, Highest Utilization
Medicare Part D Brand-Name Drugs, by Therapeutic Class, 2016

Therapeutic class                                    Drugs             Expenditures        Rebates and other Rebates and other
                                                                         (in dollars)      price concessions price concessions
                                                                                                   (in dollars) as proportion of
                                                                                                                   expenditures
                                                                                                                       (percent)
Endocrine metabolic agent                                69           21,159,397,592             8,706,047,671                     41
Anti-infective agent                                     34           11,225,406,324             2,217,176,860                     20
Respiratory agent                                        31            9,886,018,944             3,175,329,544                     32
Central nervous system agent                             57            8,505,550,786             1,910,305,516                     22
Antineoplastic agent                                     81            8,371,964,875               169,735,568                       2
Cardiovascular agent                                     51            8,002,376,936             3,064,718,411                     38
Immunological agent                                      28            6,307,238,469               256,298,135                       4
Blood modifier agent                                     19            5,392,226,626             1,444,676,699                     27
Genitourinary agent                                      14            3,561,199,830             1,406,975,264                     40
Musculoskeletal agent                                    10            3,515,156,333               589,240,842                     17


                                         4
                                          Part D plans are required to provide access to all drugs covered under certain
                                         therapeutic classes of drugs, known as Medicare protected classes: (1) anticonvulsants,
                                         (2) antidepressants, (3) antineoplastics, (4) antipsychotics, (5) antiretrovirals, and (6)
                                         immunosuppressants for the treatment of transplant. Antineoplastic and immunological
                                         drugs, which are part of Medicare’s protected classes, received the lowest amount of
                                         direct and indirect remuneration (DIR) relative to their gross Part D expenditures. Because
                                         the therapeutic class information in Truven Health Analytics’ Redbook does not align with
                                         Medicare’s protected classes, we were unable to determine which of the 444 highest
                                         expenditure, highest utilization drugs fell in under the protected classes. Furthermore, it is
                                         unclear if antineoplastic and immunological drugs received relatively few rebates and
                                         other price concessions as result of the protected class requirement or other factors.




                                         Page 51                                                          GAO-19-498 Medicare Part D
                                                                Appendix IV: Expenditures and Rebate and
                                                                Other Price Concession Information for
                                                                Medicare Part D Drugs




Therapeutic class                                                               Drugs                     Expenditures            Rebates and other Rebates and other
                                                                                                            (in dollars)          price concessions price concessions
                                                                                                                                          (in dollars) as proportion of
                                                                                                                                                          expenditures
                                                                                                                                                              (percent)
Gastrointestinal agent                                                                17                 3,077,243,117                1,430,029,308                  46
Ophthalmologic agent                                                                  14                 2,930,528,225                1,332,099,094                  45
Source: GAO analysis of Centers for Medicare & Medicaid Services (CMS) data and data from Truven Health Analytics. | GAO-19-498

                                                                Notes: We analyzed CMS expenditure and rebate and other price concession information, known as
                                                                direct and indirect remuneration, for the following groups of brand-name drugs: the 200 drugs with the
                                                                highest expenditures, the 200 drugs with the highest utilization, and the 200 drugs with the highest
                                                                expenditures per 30-day prescription. We identified drugs that had common ingredients, strengths,
                                                                dose, and route of administration. This resulted in a group of 444 unique brand-name drugs across
                                                                the three groups of drugs. We identified therapeutic class information using information from Red
                                                                Book, a compendium published by Truven Health Analytics.
                                                                We omitted the six therapeutic classes in our analysis with fewer than 10 drugs in their class: nasal
                                                                agents, dermatological agents, antidotes, nutritive agents, dependency agents, and diagnostic
                                                                agents.


Gross and Net Part D                                            Rebates and other price concessions as a proportion of gross
Expenditures Varied                                             expenditures varied from 4 percent to 27 percent in 2016 for the 17 Part
                                                                D plan sponsors whose representatives we interviewed. Gross Part D
among Selected Part D
                                                                expenditures per beneficiary ranged from $1,772 to $5,583, and net Part
Plan Sponsors in 2016                                           D expenditures per beneficiary ranged from $1,687 to $4,837 (see table
                                                                6).

Table 6: Gross and Net Expenditure Information for the 17 Medicare Part D Plan Sponsors GAO Interviewed, 2016

Part D Sponsor                                     2016 Rebates and other price                                      2016 Gross Part D                 2016 Net Part D
                                           concessions as a percentage of gross                                           expenditures                    expenditures
                                                  Part D expenditures (percent)                                per Beneficiary (dollars)       per beneficiary (dollars)
Sponsor A                                                                                         27                                  4,609                       3,359
Sponsor B                                                                                         27                                  4,262                       3,125
Sponsor C                                                                                         25                                  3,640                       2,728
Sponsor D                                                                                         18                                  3,412                       2,801
Sponsor E                                                                                         17                                  2,800                       2,312
Sponsor F                                                                                         17                                  3,350                       2,777
Sponsor G                                                                                         17                                  3,031                       2,529
Sponsor H                                                                                         16                                  4,132                       3,468
Sponsor I                                                                                         16                                  5,503                       4,618
Sponsor J                                                                                         16                                  4,398                       3,709
Sponsor K                                                                                         13                                  2,753                       2,385
Sponsor L                                                                                         13                                  5,583                       4,837
Sponsor M                                                                                         11                                  3,379                       3,004




                                                                Page 52                                                                       GAO-19-498 Medicare Part D
                                                               Appendix IV: Expenditures and Rebate and
                                                               Other Price Concession Information for
                                                               Medicare Part D Drugs




Part D Sponsor                                     2016 Rebates and other price                           2016 Gross Part D                   2016 Net Part D
                                           concessions as a percentage of gross                                expenditures                      expenditures
                                                  Part D expenditures (percent)                     per Beneficiary (dollars)         per beneficiary (dollars)
Sponsor N                                                                                   11                             4,521                             4,039
Sponsor O                                                                                   10                             2,597                             2,347
Sponsor P                                                                                    5                             1,772                             1,687
Sponsor Q                                                                                    4                             2,806                             2,700
Source: GAO analysis of Centers for Medicare & Medicaid Services (CMS) data. | GAO-19-498

                                                               Notes: We analyzed CMS expenditure and rebate and other price concession information, known as
                                                               direct and indirect remuneration, for the following groups of brand-name drugs: the 200 with the
                                                               highest expenditures, the 200 with the highest utilization, and the 200 with the highest expenditures
                                                               per 30-day prescription. We identified drugs that had common ingredients, strengths, dose, and route
                                                               of administration. This resulted in a group of 444 unique brand-name drugs across the three groups
                                                               of drugs.
                                                               Gross expenditures reflect what was paid to the pharmacy by the Part D plan sponsor, pharmacy
                                                               benefit managers on the sponsor’s behalf, and the beneficiary for a given drug. Net expenditures
                                                               reflect any rebates and discounts obtained by plan sponsors and pharmacy benefit managers after a
                                                               beneficiary receives a drug.




                                                               Page 53                                                               GAO-19-498 Medicare Part D
Appendix V: Information on Discounts Off
                            Appendix V: Information on Discounts Off
                            Manufacturer List Prices for Brand-Name and
                            Generic Medicare Part D Drugs


Manufacturer List Prices for Brand-Name
and Generic Medicare Part D Drugs
                            This appendix contains additional information on the gross and net
                            discounts for the highest expenditure, highest utilization brand-name and
                            generic Medicare Part D drugs in 2016.


Information on the Extent   The amount of discounts in 2016 for the 444 highest expenditure, highest
to Which Brand-Name         utilization brand-name drugs varied by whether they were sold in retail or
                            specialty pharmacies. Discounts also varied by whether the brand-name
Part D Drugs Were
                            drugs were highest expenditure, highest utilization or highest expenditure
Discounted Off              per utilization drugs. 1 Of the 444 highest expenditure, highest utilization
Manufacturer List Prices    brand-name drugs, 244 were sold in retail pharmacies and 200 were sold
                            in specialty pharmacies. 2

                            •   Brand-name retail drugs. The three groups of drugs all had
                                pharmacy acquisition costs that were 81 percent of manufacturer list
                                prices and gross Part D prices that were between 83 and 84 percent
                                of manufacturer list prices in 2016. However, the net prices varied,
                                ranging from 55 percent of manufacturer list price for the highest
                                utilization drugs to 77 percent for the highest expenditure per
                                utilization drugs (see table 7). 3
                            •   Brand-name specialty drugs. The 38 highest expenditure drugs and
                                187 highest expenditure per utilization drugs sold in specialty
                                pharmacies had median gross prices that were between 84 and 85

                            1
                             Prices are the 2016 median prices for the brand-name drugs that met the following
                            criteria: the 200 drugs with the highest expenditures, the 200 drugs with the highest
                            utilization, and the 200 drugs with the highest expenditures per 30-day prescription. We
                            identified drugs that had common ingredients, strengths, dosages, or routes of
                            administration. This resulted in a list of 444 unique brand-name drugs across the three
                            groups because some drugs met more than one criterion and therefore appeared in more
                            than one group.
                            2
                             We determined whether a drug was sold in retail or specialty pharmacies based on
                            National Drug Acquisition Cost (NADAC data). We found that 97 percent of the 444
                            highest expenditure, highest utilization brand-name drugs that lacked NADAC were listed
                            as specialty drugs on a Part D plan sponsor’s formulary. Given this, we refer to drugs that
                            do not have a NADAC price as those sold in specialty pharmacies.
                            3
                             Manufacturer list price is the median average wholesale price—the list price
                            manufacturers suggest wholesalers charge pharmacies for a drug. Pharmacy acquisition
                            cost reflects prices reported in surveys of community retail pharmacies in the NADAC data
                            set. Gross Part D prices reflect median unit prices paid to pharmacies by Part D plan
                            sponsors, pharmacy benefit managers (PBM) on the sponsor’s behalf, and the
                            beneficiary, and net Part D prices account for rebates and other price concessions
                            obtained by plan sponsors for these drugs as reported in the direct and indirect
                            remuneration (DIR) data set.




                            Page 54                                                        GAO-19-498 Medicare Part D
                                                                Appendix V: Information on Discounts Off
                                                                Manufacturer List Prices for Brand-Name and
                                                                Generic Medicare Part D Drugs




                                                                      percent of manufacturer list price and net prices that were 84 percent
                                                                      of manufacturer list price in 2016. 4

Table 7: Medicare Part D Median Unit Drug Part D Prices as a Percentage of Manufacturer List Prices for the Highest
Expenditure, Highest Utilization Brand-Name Drugs Sold in Retail and Specialty Pharmacies, 2016

                                  Brand-name retail drugs                                                             Brand-name specialty drugs
                        Count              Pharmacy Gross Part D                          Net Part D              Count           Pharmacy Gross Part D      Net Part D
                      (number             acquisition      price                               price          (number of         acquisition      price           price
                     of drugs)          cost (percent) (percent)                           (percent)              drugs)               cost   (percent)       (percent)
                                                                                                                                            a
                                                                                                                                  (percent)
 Median 200                  162                     81                     83                     59                      38          n/a            85              84
 highest
 expenditure
 Median 200                  199                     81                     83                     55                        1         n/a           n/ab            n/ab
 highest
 utilization
 Median 200                   13                     81                     84                     77                      187         n/a            84              84
 highest
 expenditure
 per
 utilization
Legend: n/a = not applicable
Source: GAO analysis of Centers for Medicare & Medicaid Services (CMS) data and data from Truven Analytics. I GAO-19-498

                                                                Notes: Prices are the 2016 median prices for the brand-name drugs that met these criteria: the 200
                                                                drugs with the highest expenditures, the 200 drugs with the highest utilization, and the 200 drugs with
                                                                the highest expenditures per utilization (i.e., highest expenditure per number of 30-day prescriptions).
                                                                We identified drugs that had common ingredients, strengths, dosages, and routes of administration.
                                                                This resulted in a group of 444 unique brand-name drugs across the three groups because some
                                                                drugs met multiple criteria and appeared in more than one group. Of these 444 brand-name drugs,
                                                                244 were drugs sold in retail pharmacies, and 200 were drugs sold in specialty pharmacies.
                                                                Manufacturer list price is the median average wholesale price. Pharmacy acquisition cost reflects
                                                                prices reported in surveys of community retail pharmacies in the National Average Drug Acquisition
                                                                Cost data set. Gross Part D prices reflect median unit prices paid to pharmacies by Part D plan
                                                                sponsors, pharmacy benefit managers on the sponsor’s behalf, and the beneficiary and net Part D
                                                                prices account for rebates and other price concessions obtained by plan sponsors for these drugs as
                                                                reported in the direct and indirect remuneration data set.
                                                                a
                                                                 Pharmacy acquisition cost data are unavailable for drugs sold in specialty pharmacies, as these
                                                                pharmacies are not surveyed by CMS.
                                                                b
                                                                 Only one of the highest utilization drugs was sold in specialty pharmacies; therefore, we omitted
                                                                pricing data on that drug.


                                                                We also found variation in brand-name prices across therapeutic classes
                                                                for the 244 highest expenditure, highest utilization Part D drugs sold in

                                                                4
                                                                 Pharmacy acquisition cost data are unavailable for drugs sold in specialty pharmacies, as
                                                                non-retail pharmacies are not surveyed by the Centers for Medicare & Medicaid Services
                                                                (CMS). Only one of the highest utilization drugs was sold in specialty pharmacies;
                                                                therefore, we omitted pricing data on that drug.




                                                                Page 55                                                                      GAO-19-498 Medicare Part D
                                                                Appendix V: Information on Discounts Off
                                                                Manufacturer List Prices for Brand-Name and
                                                                Generic Medicare Part D Drugs




                                                                retail pharmacies. 5 In 2016, median gross Part D prices for the brand-
                                                                name drugs sold in retail pharmacies were similar across the nine
                                                                therapeutic classes we analyzed, ranging from 81 percent to 84 percent
                                                                of the manufacturer list price. 6 However, there was a much wider range
                                                                among median net prices, from 43 percent to 83 percent of manufacturer
                                                                list price. Anti-infective agents had the lowest percentage point changes
                                                                in their prices from gross to net (1 percentage point), while endocrine
                                                                metabolic agents, cardiovascular agents, respiratory agents,
                                                                ophthalmologic agents, and genitourinary agents had the largest
                                                                changes, with declines from gross to net of greater than 30 or more
                                                                percentage points (see table 8).

Table 8: Medicare Part D Median Unit Drug Prices as a Percentage of Manufacturer List Prices by Therapeutic Class for the
Highest Expenditure, Highest Utilization Brand-name Drugs Sold in Retail Pharmacies, 2016

Therapeutic Class                                                      Count              Pharmacy acquisition                    Gross Part D price    Net Part D price
                                                                                                cost (percent)                            (percent)           (percent)
Endocrine metabolic agent                                                    48                                        81                        82                  52
Central nervous system agent                                                 39                                        80                        84                  60
Cardiovascular agent                                                         27                                        81                        81                  50
Respiratory agent                                                            26                                        81                        83                  51
Anti-infective agent                                                         22                                        81                        84                  83
Ophthalmologic agent                                                         14                                        81                        83                  43
Gastrointestinal agent                                                       13                                        80                        84                  56
Genitourinary agent                                                          13                                        81                        83                  44
Blood modifier agent                                                         12                                        81                        83                  60
Source: GAO analysis of Centers for Medicare & Medicaid Services (CMS) data and data from Truven Health Analytics. | GAO-19-498

                                                                Notes: Prices are the 2016 median prices for the 244 brand-name drugs that were sold in retail
                                                                pharmacies and were among the following three groups of drugs which included drugs sold in retail
                                                                and specialty pharmacies: the 200 drugs with the highest expenditures, the 200 drugs with the
                                                                highest utilization, and the 200 drugs with the highest expenditures per 30-day prescription. We
                                                                identified drugs that had common ingredients, strengths, dosages, and routes of administration. This
                                                                resulted in a group of 444 unique brand-name drugs across the three groups because some drugs
                                                                met multiple criteria and appeared in more than one group. Of these 444 brand-name drugs 244 were
                                                                drugs sold in retail pharmacies, and 200 were drugs sold in specialty pharmacies.

                                                                5
                                                                 Of the 244 highest expenditure, highest utilization brand-name drugs sold in retail
                                                                pharmacies, 19.7 percent were endocrine metabolic agents; 16 percent were central
                                                                nervous system agents; 11.1 percent were cardiovascular agents; and 10.7 percent were
                                                                respiratory agents.
                                                                6
                                                                 We omitted the eight therapeutic classes in our analysis with fewer than 10 drugs in their
                                                                class: antidotes, antineoplastic agents, dermatological agents, dependency agents,
                                                                immunological agents, musculoskeletal agents, nasal agents, and nutritive agents; these
                                                                accounted for 30 brand-name drugs sold in retail pharmacies.




                                                                Page 56                                                                        GAO-19-498 Medicare Part D
                            Appendix V: Information on Discounts Off
                            Manufacturer List Prices for Brand-Name and
                            Generic Medicare Part D Drugs




                            We omitted the eight therapeutic classes in our analysis with fewer than 10 drugs in their class:
                            antidotes, antineoplastic agents, dermatological agents, dependency agents, immunological agents,
                            musculoskeletal agents, nasal agents, and nutritive agents; these accounted for 30 brand-name
                            drugs sold in retail pharmacies. We identified therapeutic class information using information from
                            Red Book, a compendium published by Truven Health Analytics.

                            Manufacturer list price is the median average wholesale price—the list price manufacturers suggest
                            wholesalers charge pharmacies for a drug. Pharmacy acquisition cost reflects prices reported in
                            surveys of community retail pharmacies in the National Average Drug Acquisition Cost data set.
                            Gross Part D prices reflect median unit prices paid to pharmacies by Part D plan sponsors, pharmacy
                            benefit managers on the sponsor’s behalf, and the beneficiary and net Part D prices account for
                            rebates and other price concessions obtained by plan sponsors for these drugs as reported in the
                            direct and indirect remuneration data set.


                            In contrast, there was little variation in both median gross and net prices
                            across all therapeutic classes for brand-name drugs sold in specialty
                            pharmacies. The range in median gross prices as a proportion of
                            manufacturer list prices across the six therapeutic classes was 83 percent
                            to 86 percent, and the range in median net prices as a proportion of
                            manufacturer list prices was 80 percent to 84 percent. 7


Information on the Extent   In 2016, discounts off of the manufacturer list price varied by whether the
to Which Generic Part D     generic drug was sold in retail pharmacies or in specialty pharmacies. Of
                            the 476 highest expenditure, highest utilization generic drugs in our
Drugs Were Discounted
                            analysis, the 367 sold in retail pharmacies had a median gross and net
Off Manufacturer List       Part D price that were 66 percentage points lower than the manufacturer
Prices                      list price, and 13 percentage points higher than the pharmacy’s cost of
                            acquiring the drugs. 8

                            The 109 generic drugs sold in specialty pharmacies received far fewer
                            discounts off of manufacturer list price than drugs sold in retail
                            7
                             Four therapeutic classes accounted for approximately 71 percent of the 200 brand-name
                            drugs sold in specialty pharmacies: 39 percent were antineoplastic agents; 12 percent
                            were cardiovascular agents; 11 percent were immunological agents; and 11 percent were
                            endocrine metabolic agents. Anti-infective agents, blood modifier agents, central nervous
                            system agents, and respiratory agents each accounted for less than 10 percent of the 200
                            drugs. We omitted one drug without a therapeutic class identified and the eight therapeutic
                            classes with fewer than 10 drugs in the class from our analysis: blood modifier agents,
                            dermatologic agents, diagnostic agents, gastrointestinal agents, genitourinary agents,
                            musculoskeletal agents, respiratory agents, and antidotes.
                            8
                              Prices are the 2016 median prices for the generic drugs that met these criteria: the 200
                            drugs with the highest expenditures, the 200 drugs with the highest utilization, and the 200
                            drugs with the highest expenditures per 30-day prescription. We identified drugs that had
                            common ingredients, strengths, dosages, and routes of administration. This resulted in a
                            list of 476 unique brand-name drugs across the three groups, drugs some drugs met more
                            than one criterion and, therefore, appeared in more than one group.




                            Page 57                                                              GAO-19-498 Medicare Part D
Appendix V: Information on Discounts Off
Manufacturer List Prices for Brand-Name and
Generic Medicare Part D Drugs




pharmacies. Median gross and net prices for those drugs sold in specialty
pharmacies were both 26 percentage points lower than manufacturer list
prices (see fig. 8). Therefore, generic drugs sold in retail pharmacies
received median discounts (66 percent below manufacturer list prices)
that were 2.5 times larger than those generic drugs sold in specialty
pharmacies (26 percent below manufacturer list prices).

Figure 8: Medicare Part D Drug Prices as a Percentage of Manufacturer List Prices
for the Highest Expenditure, Highest Utilization Generic Drugs Sold in Retail and
Specialty Pharmacies, 2016




Notes: Generic drugs sold in retail pharmacies had a gross median per unit price of $0.64, while the
gross median per unit price for generic drugs sold in specialty pharmacies was $22.66.
Prices are the 2016 median prices for the generic drugs that met the following criteria: the 200 drugs
with the highest expenditures, the 200 drugs with the highest utilization, and the 200 drugs with the
highest expenditures per 30-day prescription. We identified drugs that had common ingredients,
strengths, dose, and route of administration. This resulted in a group of 476 unique generic drugs
across the three groups because some drugs met multiple criteria and appeared in more than one
group. Of these 476 generic drugs, 367 were drugs sold in retail pharmacies and 109 were drugs sold
in specialty pharmacies.
Manufacturer list price is the median average wholesale price. Pharmacy acquisition cost reflects
prices reported in surveys of community retail pharmacies in the National Average Drug Acquisition



Page 58                                                               GAO-19-498 Medicare Part D
Appendix V: Information on Discounts Off
Manufacturer List Prices for Brand-Name and
Generic Medicare Part D Drugs




Cost data set. Gross Part D prices reflect median unit prices paid to pharmacies by Part D plan
sponsors, pharmacy benefit managers on the sponsor’s behalf, and the beneficiary and net Part D
prices account for rebates and other price concessions obtained by plan sponsors for these drugs as
reported in the direct and indirect remuneration data set.
a
 Pharmacy acquisition cost data are unavailable for drugs sold in specialty pharmacies, as these
pharmacies are not surveyed by CMS.


We also found pricing variation by whether the generic drugs were in the
200 highest expenditure, 200 highest utilization group, or the 200 highest
expenditures per utilization group.

•    Generic retail drugs. Of the 367 generic drugs sold in retail
     pharmacies, 200 were in the group of the 200 highest utilization
     generic drugs, 198 were in the group of the 200 highest expenditure
     generic drugs, and 91 were in the group of the 200 generic drugs with
     the highest expenditure per utilization. 9 We found that the gross Part
     D price for the highest utilization drugs was 14 percent of the
     manufacturer list price, while the gross price for the highest
     expenditure drugs was 34 percent of the manufacturer list price.
     However, the Part D gross price for the highest expenditure per
     utilization drugs was 63 percent of the manufacturer list price. The
     difference in gross and net Part D price as a percentage of
     manufacturer list price was one percentage point or less for all three
     groups of drugs (see table 9).
•    Generic specialty drugs. Of the 109 generic drugs sold in specialty
     pharmacies, none was in the group of the 200 highest utilization
     generic drugs, two were in the group of the 200 highest expenditure
     generic drugs, and all 109 were in the group of the 200 highest
     expenditure per utilization generic drugs. 10 The gross Part D price for
     the highest expenditure per utilization drugs sold in specialty
     pharmacies was 74 percent of the manufacturer list price, and these
     drugs received no additional rebates and other price concessions.




9
 Drugs may be included in more than one of our three groups. For example, some drugs
are included in both the group of 200 highest expenditure brand-name drugs sold in retail
pharmacies and the group of 200 highest expenditure per utilization drugs.
10
  The two generic drugs sold in specialty pharmacies that were in the group of 200
highest expenditure generic drugs were also in the group of the 200 highest expenditure
per utilization drugs.




Page 59                                                               GAO-19-498 Medicare Part D
                                                                Appendix V: Information on Discounts Off
                                                                Manufacturer List Prices for Brand-Name and
                                                                Generic Medicare Part D Drugs




Table 9: Medicare Part D Median Unit Drug Prices as a Percentage of Manufacturer List Prices for the Highest Expenditure,
Highest Utilization Generic Drugs Sold in Retail and Specialty Pharmacies, 2016

                  Generic retail drugs as a percentage of manufacturer                                 Generic specialty drugs as a percentage of manufacturer
                                        list price                                                                             list price
                      Count                  Pharmacy                  Gross Net Part D                        Count                   Pharmacy Gross Part D       Net Part D
                  (number of            acquisition cost               Part D     Price                    (number of             acquisition cost     price            price
                      drugs)                   (percent)                price (percent)                        drugs)                   (percent)a (percent)        (percent)
                                                                    (percent)
 Median 200                  198                            20                34                33                        2                  n/a            n/ab         n/ab
 highest
 expenditure
 Median 200                  200                              6               14                14                        0                  n/a            n/ab         n/ab
 highest
 utilization
 Median 200                    91                           60                63                62                    109                    n/a              74          74
 highest
 expenditure
 per
 utilization
Legend: n/a = not applicable
Source: GAO analysis of Centers for Medicare & Medicaid Services (CMS) data and data from Truven Health Analytics. | GAO-19-498

                                                                Notes: Prices are the 2016 median prices for the generic drugs that met these criteria: the 200 drugs
                                                                with the highest expenditures, the 200 drugs with the highest utilization, and the 200 drugs with the
                                                                highest expenditures per 30-day prescription. We identified drugs that had common ingredients,
                                                                strengths, dose, and route of administration. This resulted in a group of 476 unique generic drugs
                                                                across the three groups because some drugs met multiple criteria and appeared in more than one
                                                                group. Of these 476 generic drugs, 367 were drugs sold in retail pharmacies and 109 were drugs sold
                                                                in specialty pharmacies.
                                                                Manufacturer list price is the median average wholesale price. Pharmacy acquisition cost reflects
                                                                prices reported in surveys of community retail pharmacies in the National Average Drug Acquisition
                                                                Cost data set. Gross Part D prices reflect median unit prices paid to pharmacies by Part D plan
                                                                sponsors, pharmacy benefit managers on the sponsor’s behalf, and the beneficiary and net Part D
                                                                prices account for rebates and other price concessions obtained by plan sponsors for these drugs as
                                                                reported in the direct and indirect remuneration data set.
                                                                a
                                                                 Pharmacy acquisition cost data are unavailable for drugs sold in specialty pharmacies, as these
                                                                pharmacies are not surveyed by CMS.
                                                                b
                                                                Only two of the highest expenditure generic drugs and none of the highest utilization generic drugs
                                                                were sold in specialty pharmacies; therefore, we omitted pricing data for these drugs.


                                                                There was variation in generic drug pricing across the eight therapeutic
                                                                classes for generic drugs sold in retail pharmacies. Median gross Part D
                                                                prices for generic retail drugs ranged from 14 percent of manufacturer list
                                                                prices for cardiovascular agents to 56 percent of manufacturer list prices
                                                                for dermatological agents (see table 10). However, there was little
                                                                difference between in median gross and net Part D prices as a
                                                                percentage of manufacturer list price for generic retail drugs in any




                                                                Page 60                                                                            GAO-19-498 Medicare Part D
                                                                Appendix V: Information on Discounts Off
                                                                Manufacturer List Prices for Brand-Name and
                                                                Generic Medicare Part D Drugs




                                                                therapeutic class, with the percentage difference ranging from 0 percent
                                                                to 2 percent. 11

Table 10: Medicare Part D Median Unit Drug Prices as a Percentage of Manufacturer List Prices by Therapeutic Class for the
Highest Expenditure, Highest Utilization Generic Drugs Sold in Retail Pharmacies, 2016

Therapeutic class                                                      Count             Pharmacy acquisition                  Gross Part D price   Net Part D price
                                                                                               cost (percent)                          (percent)          (percent)
Central nervous system agent                                               127                                         17                     30                   29
Cardiovascular agent                                                         88                                            6                  14                   14
Endocrine metabolic agent                                                    33                                        46                     54                   52
Anti-infective agent                                                         22                                        43                     51                   51
Gastrointestinal agent                                                       20                                        32                     46                   45
Dermatological agent                                                         18                                        49                     56                   55
Blood modifier agent                                                         15                                        25                     38                   38
Genitourinary agent                                                          10                                        30                     40                   39
Source: GAO analysis of Centers for Medicare & Medicaid Services (CMS) data and data from Truven Analytics. I GAO-19-498

                                                                Notes: Prices are the 2016 median prices for the 367 generic drugs that were sold in retail
                                                                pharmacies and were among the following three groups of drugs which included drugs sold in retail
                                                                and specialty pharmacies: the 200 drugs with the highest expenditures, the 200 drugs with the
                                                                highest utilization, and the 200 drugs with the highest expenditures per 30-day prescription. We
                                                                identified drugs that had common ingredients, strengths, dose, and route of administration. This
                                                                resulted in a group of 476 unique generic drugs across the three groups because some drugs met
                                                                multiple criteria and appeared in more than one group. Of these 476 generic drugs, 367 were drugs
                                                                sold in retail pharmacies and 109 were drugs sold in specialty pharmacies.
                                                                We omitted one drug without therapeutic class information and the eight therapeutic classes in our
                                                                analysis with fewer than 10 drugs in their class: antineoplastic agents, antidotes, immunologic agents,
                                                                musculoskeletal agents, nasal agents, nutritive agents, ophthalmologic agents, and respiratory
                                                                agents; these accounted for 33 generic drugs sold in retail pharmacies. We identified therapeutic
                                                                class information using information from Red Book, a compendium published by Truven Health
                                                                Analytics.
                                                                Manufacturer list price is the median average wholesale price—the list price manufacturers suggest
                                                                wholesalers charge pharmacies for a drug. Pharmacy acquisition cost reflects prices reported in
                                                                surveys of community retail pharmacies in the National Average Drug Acquisition Cost data set.
                                                                Gross Part D prices reflect median unit prices paid to pharmacies by Part D plan sponsors, pharmacy
                                                                benefit managers on the sponsor’s behalf, and the beneficiary and net Part D prices account for
                                                                rebates and other price concessions obtained by plan sponsors for these drugs as reported in the
                                                                direct and indirect remuneration data set.


                                                                There was little variation in median gross and net prices across the
                                                                therapeutic classes for generic drugs sold in specialty pharmacies. The

                                                                11
                                                                  We omitted one drug without therapeutic class information and the eight therapeutic
                                                                classes in our analysis with fewer than 10 drugs in their class: antineoplastic agents,
                                                                antidotes, immunologic agents, musculoskeletal agents, nasal agents, nutritive agents,
                                                                ophthalmologic agents, respiratory agents; these accounted for 33 generic drugs sold in
                                                                retail pharmacies.




                                                                Page 61                                                                    GAO-19-498 Medicare Part D
                                                                Appendix V: Information on Discounts Off
                                                                Manufacturer List Prices for Brand-Name and
                                                                Generic Medicare Part D Drugs




                                                                range in median gross prices as a percentage of manufacturer list prices
                                                                was 73 to 75 percent (see table 11). There was little difference between
                                                                median gross and Part D net prices as a percentage of manufacturer list
                                                                price, with the percentage difference between median gross and net
                                                                prices 1 percent or less for all classes. 12

Table 11: Medicare Part D Median Unit Drug Prices as a Percentage of Manufacturer List Prices by Therapeutic Class for the
Highest Expenditure, Highest Utilization Generic Drugs Sold in Specialty Pharmacies, 2016

Therapeutic class                                                              Count                   Pharmacy             Gross Part D price     Net Part D price
                                                                                                 acquisition costa                  (percent)            (percent)
Anti-infective agent                                                                42                                n/a                   73                      72
Central nervous system agent                                                        27                                n/a                   75                      75
Antineoplastic agent                                                                16                                n/a                   74                      74
Legend: n/a = not applicable
Source: GAO analysis of Centers for Medicare & Medicaid Services (CMS) data and data from Truven Analytics. I GAO-19-498

                                                                Notes: Prices are the 2016 median prices for the 109 generic drugs that were sold in specialty
                                                                pharmacies and were among the following three groups of drugs which included drugs sold in retail
                                                                and specialty pharmacies: the 200 drugs with the highest expenditures, the 200 drugs with the
                                                                highest utilization, and the 200 drugs with the highest expenditures per 30-day prescription. We
                                                                identified drugs that had common ingredient, strengths, dose, and route of administration. This
                                                                resulted in a group of 476 unique generic drugs across the three groups because some drugs met
                                                                multiple criteria and appeared in more than one group. Of these 476 generic drugs, 367 were drugs
                                                                sold in retail pharmacies and 109 were drugs sold in specialty pharmacies.
                                                                We omitted the 9 therapeutic classes in our analysis with fewer than 10 drugs in their class: antidotes,
                                                                blood modifier agents, cardiovascular agents, dermatological agents, endocrine metabolic agents,
                                                                gastrointestinal agents, immunological agents, nutritive agents, and respiratory agents; these
                                                                accounted for 24 generic drugs sold in specialty pharmacies. We identified therapeutic class
                                                                information using information from Red Book, a compendium published by Truven Health Analytics.
                                                                Manufacturer list price is the median average wholesale price—the list price manufacturers suggest
                                                                wholesalers charge pharmacies for a drug. Pharmacy acquisition cost reflects prices reported in
                                                                surveys of community retail pharmacies in the National Average Drug Acquisition Cost data set.
                                                                Gross Part D prices reflect median unit prices paid to pharmacies by Part D plan sponsors, pharmacy
                                                                benefit managers on the sponsor’s behalf, and the beneficiary and net Part D prices account for
                                                                rebates and other price concessions obtained by plan sponsors for these drugs as reported in the
                                                                direct and indirect remuneration data set.
                                                                a
                                                                 Pharmacy acquisition cost data are unavailable for drugs sold in specialty pharmacies, as they are
                                                                not surveyed by CMS.




                                                                12
                                                                  We omitted the nine therapeutic classes in our analysis with fewer than 10 drugs in their
                                                                class: antidotes, blood modifier agents, cardiovascular agents, dermatological agents,
                                                                endocrine metabolic agents, gastrointestinal agents, immunological agents, nutritive
                                                                agents, and respiratory agents; these accounted for 24 generic drugs sold in specialty
                                                                pharmacies.




                                                                Page 62                                                                 GAO-19-498 Medicare Part D
Appendix VI: Studies and Stakeholders’
                         Appendix VI: Studies and Stakeholders’ Views
                         on Effects of Utilization Management Services



Views on Effects of Utilization Management
Services
                         This appendix contains additional details on our review of 52 peer-
                         reviewed studies on the effects of utilization management services on (1)
                         financial savings, (2) beneficiary health indicators, and (3) beneficiary
                         medication adherence and access, as well as stakeholders’ views on
                         these effects.


Effect of Utilization    Of the 36 studies that examined the effect of utilization management
Management Services on   services on financial savings, 18 examined medication therapy
                         management programs and eight examined generic substitution. 1 The
Financial Savings
                         two groups of studies found the following:

                         •   Medication therapy management programs or comprehensive
                             medical reviews. Thirteen of the 18 studies that examined the
                             relationship between a medication therapy management program or
                             comprehensive medical review and financial savings found an
                             increase in savings. 2 For example, one study found that a medication
                             therapy management program conducted by telephone decreased
                             beneficiary drug costs by $682 per beneficiary for participants,
                             compared to an increase of $119 for those not in the program. 3
                         •   Generic and therapeutic substitution and generic dispensing
                             rate. Of the 8 studies that examined the relationship between generic
                             and therapeutic substitution and financial savings, all found an


                         1
                          The other 10 studies examined the effect of various combinations of step therapy, prior
                         authorization and medicine quantity limits on financial savings. Six of these studies found
                         an increase in financial savings and four found no impact, a mixed impact, or a decrease
                         in financial savings.
                         2
                          The Centers for Medicare & Medicaid Services (CMS) requires medication therapy
                         management programs for covered drugs furnished through a Part D plan. These
                         programs are generally designed to reduce the risk of adverse drug events through
                         discussion with targeted beneficiaries and prescriber intervention. A comprehensive
                         medical review is a systematic process of assessing medications to identify problems,
                         such as a change in a drug’s effect when taken with another drug, and creating a plan to
                         resolve them.
                         3
                          L.R. Moczygemba, J. C. Barner, J. C. Brannier, and E. R. Gabrillo, “Outcomes of a
                         Medicare Part D Telephone Medication Therapy Management Program,” Journal of the
                         American Pharmacists Association, vol. 52, no. 6 (2012): e144-e152. Nonparticipants in
                         the program were matched to medication therapy management program participants by
                         the beneficiary’s number of chronic diseases and Part D drugs. Because this study reports
                         outcomes from a regional Part D telephone medication therapy management program, the
                         results may not be generalizable to all Part D plans or face-to-face medication therapy
                         management programs.




                         Page 63                                                         GAO-19-498 Medicare Part D
Appendix VI: Studies and Stakeholders’ Views
on Effects of Utilization Management Services




    increase in savings. 4 For example, a 2013 study examined the
    potential financial savings to beneficiaries and Part D plan sponsors of
    generic and therapeutic substitution of commonly prescribed drugs. 5
    The study estimated that in 2007, generic and therapeutic
    substitutions could have resulted in an average annual savings of
    $127 and $389 per person, respectively. 6

Additionally, eight of these 36 studies examined the generic dispensing
rate, and all eight found that utilization management led to an increase in
the rate. 7 The generic dispensing rate—the percent of prescriptions
dispensed with a generic drug instead of a brand-name drug—represents
a source of financial savings through a reduction in the use of brand-
name drugs, which are generally more expensive than generics. For
example, a 2017 study analyzed 2012 Part D data to examine the impact
of prior authorization and step therapy on generic use among low-income
subsidy beneficiaries. 8 This study found that those randomly assigned to

4
 Studies generally calculated potential savings by estimating the difference between the
amount paid for brand-name drugs with the amount that would have been paid if generic
drugs had been used.
5
 O.K. Duru, S. L. Ettner, N. Turk, C. M. Mangione, A. F. Brown, J. Fu, L. Simien, and C.W.
Tseng, “Potential Savings Associated with Drug Substitution in Medicare Part D: The
Translating Research Into Action for Diabetes (TRIAD) Study,” Journal of General Internal
Medicine, vol. 29, no. 1 (2014): 230-236. Generic substitution is switching a generic drug
for its bioequivalent, brand-name counterpart. Therapeutic substitution is switching a
generic drug for a brand-name drug that is not bioequivalent but is within the same
therapeutic class.
6
 The authors noted this study may overestimate potential savings as it assumes that all
beneficiaries make every substitution, which in the case of therapeutic substitution is not
always medically appropriate.
7
 Of the eight studies, three looked at the effect of medication therapy management, two
looked at step therapy, and three looked at the effect of both prior authorization and step
therapy. The generic dispensing rate was generally measured as annual days of supply
for generics divided by annual total days of supply for all drugs in the class.
8
 X. Shen, B. C. Stuart, C. A. Powers, S. E. Tom, L. S. Magder, and E. M. Perfetto, “Impact
of Formulary Restrictions on Medication Use and Costs,” The American Journal of
Managed Care, vol. 23, no. 8 (2017): e265-e274. Low-income subsidy beneficiaries with
annual incomes under 135 percent of the federal poverty level, and who meet other
criteria, qualify for a 100 percent premium subsidy and reduced cost sharing, among other
cost reductions. Prior authorization is a requirement that beneficiaries obtain approval for
a drug by the pharmacy benefit manager or plan before obtaining the drug if it is to be
covered by the plan. Step therapy is a requirement where more expensive drugs are
covered only if beneficiaries try less expensive alternatives first and find them not to be
effective.




Page 64                                                         GAO-19-498 Medicare Part D
                         Appendix VI: Studies and Stakeholders’ Views
                         on Effects of Utilization Management Services




                         a plan using both prior authorization and step therapy had an increased
                         generic dispensing rate of 3 to 15 percentage points for all three classes
                         of drugs examined.


Effect of Utilization    Twelve of the 20 studies that examined beneficiary health indicators
Management Services on   found that utilization management services were associated with
                         improved indicators, while the other eight found a mixed impact, no
Beneficiary Health       impact, or a decline. 9 Examples of studies that looked at the association
Indicators               of utilization management services with beneficiary health indicators
                         include:

                         •    A study analyzing data from three Part D plan sponsors, which found
                              there was a nearly 50 percent reduction in the use of potentially
                              harmful drugs by beneficiaries 6 months after the implementation of a
                              retrospective drug utilization review program. 10
                         •    A randomized trial of medication therapy management for Part D
                              beneficiaries found a nearly 60 percent reduction in beneficiaries’ drug
                              therapy problems over time among two groups after the medication
                              therapy management intervention. 11




                         9
                          Studies used a range of health indicators, including reducing medications inappropriate
                         for older adults, improved cholesterol values, and reductions in adverse drug events.
                         10
                           C.I. Starner, S. A. Norman, R. G. Reynolds, and P. P. Gleason, “Effect of a
                         Retrospective Drug Utilization Review on Potentially Inappropriate Prescribing in the
                         Elderly,” The American Journal of Geriatric Pharmacotherapy, vol. 7, no. 1 (2009): 11-19.
                         This study did not use a control group, so the results may not be interpreted as the causal
                         effect of the utilization management service. A drug utilization review program is a
                         concurrent examination by the pharmacy benefit manager and plan sponsor of
                         prescriptions at the time of purchase by the beneficiary to assess safety considerations,
                         such as potential adverse interactions, and compliance with clinical guidelines (including
                         quantity and dose).
                         11
                           The study examined Part D beneficiaries at three academic medical center community
                         pharmacies and family medical clinics at three sites between December 2007 and January
                         2010. Examples of drug therapy problems include: incorrect dose, drug interactions, or
                         taking the wrong drug for a beneficiary’s condition. D.R. Touchette, A. L. Masica, R. J.
                         Dolor, G. T. Schumock, Y. K. Choi, Y. Kim, and S. R. Smith, “Safety-Focused Medication
                         Therapy Management: A Randomized Controlled Trial,” Journal of the American
                         Pharmacists Association, vol. 52, no. 5 (2012): 603-612.




                         Page 65                                                        GAO-19-498 Medicare Part D
                         Appendix VI: Studies and Stakeholders’ Views
                         on Effects of Utilization Management Services




Effect of Utilization    Fifteen studies examined the effect of utilization management services on
Management Services on   beneficiary medication adherence and access. Seven of the 10 studies
                         that examined the effect of either medication therapy management
Beneficiary Medication
                         programs or comprehensive medication reviews on beneficiaries’
Adherence and Access     medication adherence (taking medication as prescribed) found
                         improvement. 12 For example, a 2016 study used data from Part D and the
                         U.S. Renal Data System to examine the relationship of medication
                         therapy management eligibility with immunosuppressant drug adherence
                         12 months after beneficiaries received a kidney transplant. 13 The study
                         found that medication therapy management-eligible transplant recipients
                         were 14 percent more likely to have improved adherence than transplant
                         recipients who were not eligible. The other three studies that examined
                         medication therapy management programs or comprehensive medication
                         reviews found no statistically significant impact on adherence.

                         The effect of two other utilization management services—prior
                         authorization and step therapy—on beneficiary medication adherence
                         and access (the ability to obtain clinically indicated prescriptions) is
                         unclear, according to the studies we reviewed. 14 The two studies that
                         examined the relationship of prior authorization and step therapy with
                         adherence both found a mixed impact. For example, one study examined
                         the impact of a health plan requiring either prior authorization or step




                         12
                           Studies varied by the design of the programs and the definitions of medication
                         adherence. For example, studies had varying eligibility criteria for participation—some
                         used opt-in enrollment (eligible beneficiaries must choose to participate), while others
                         used opt-out enrollment (eligible beneficiaries are automatically enrolled unless they
                         decline to participate). Studies also used various measures of adherence, such as surveys
                         of beneficiaries who reported if they stopped their medications. Other studies measured
                         adherence using the medication possession ratio—the proportion of days’ supply obtained
                         during a time period, and considered beneficiaries above a certain proportion of days
                         covered (e.g., 75 percent) as adherent.
                         13
                           M.A. Chisholm-Burns, C. A. Spivey, E. A. Tolley, and E. K. Kaplan, “Medication Therapy
                         Management and Adherence among US Renal Transplant Recipients,” Patient Preference
                         and Adherence, vol.10 (2016): 703-709. Immunosuppressant adherence was defined as a
                         medication possession ratio of at least 80 percent.
                         14
                           Studies used various measures of access, such as the time for medications to receive
                         prior authorization approval or surveys of physicians on whether their patients were
                         unable to access needed medications. Also, some studies calculated the combined effect
                         of prior authorization and step therapy and others examined them individually.




                         Page 66                                                      GAO-19-498 Medicare Part D
Appendix VI: Studies and Stakeholders’ Views
on Effects of Utilization Management Services




therapy on medication use among dual-eligible nursing home residents. 15
The study found that some residents whose new plan required prior
authorization or step therapy for their current medication were more likely
to have gaps in medication use than those without for two of six classes
of drugs in 2006, but no gaps for any of the classes for in 2007 and
2008. 16

The two studies that examined the relationship of prior authorization and
step therapy with access found an increase in medication access
problems, but they did not focus exclusively on the Medicare population. 17
For example, one study used 2006 data from a random sample of
psychiatrists surveyed about their patients to examine the relationship of
prior authorization and step therapy with medication access problems
among dual-eligible psychiatric patients. The study found that patients in
plans with prior authorization and step therapy requirements were 2.8 and
1.8 times more likely, respectively, to have experienced medication
access problems than patients in plans without these requirements. 18 This
study examined the transition of dual-eligible beneficiaries from Medicaid
drug coverage to Medicare Part D when the program began in 2006, so
the results may not be generalizable to the entire Medicare population at
present.




15
  Dual-eligible beneficiaries are eligible for both Medicare and Medicaid. H.A. Huskamp,
D. G. Stevenson, A. J. O’Malley, S. B. Dusetzina, S. L. Mitchell, B. J. Zarowitz, M. E.
Chernew, and J. P. Newhouse, “Medicare Part D Plan Generosity and Medication Use
among Dual-Eligible Nursing Home Residents,” Medical Care, vol. 51, no. 10 (2013): 894-
900.
16
  The two classes that were more likely to have gaps in medication use for some
residents were antipsychotics and opioids. The other four classes of drugs were
angiotensin receptor blockers, cholinesterase inhibitors, osteoporosis medications, and
antidepressants.
17
  Another study that examined the relationship of only prior authorization with the time
needed to access medications found no clinically significant impact.
18
  J.C. West, J. E. Wilk, D. S. Rae, I. L. Muszynski, M. Rubio-Stipec, C. L. Alter, K. E.
Sanders, S. Crystal, and D. A. Regier, “First-Year Medicare Part D Prescription Drug
Benefits: Medication Access and Continuity among Dual Eligible Psychiatric Patients,” The
Journal of Clinical Psychiatry, vol. 71, no. 4 (2010): 400-410.




Page 67                                                         GAO-19-498 Medicare Part D
                               Appendix VI: Studies and Stakeholders’ Views
                               on Effects of Utilization Management Services




Stakeholder Perspectives       Most representatives of pharmacy benefit managers (PBM), Part D plan
on the Effect of Utilization   sponsors, and a manufacturer trade association we interviewed generally
                               agreed that utilization management services resulted in financial savings
Management on Financial
                               by requiring the use of generic drugs. Representatives of 10 of 14 plan
Savings, Beneficiary           sponsors and six of eight PBMs we interviewed stated that utilization
Health, Medication             management services generally resulted in financial savings. 19
Adherence and Access to        Representatives of one Part D plan sponsor stated that its utilization
Clinically Appropriate         management services resulted in annual savings of approximately 3
                               percent.
Medications
                               However, representatives of one Part D plan sponsor and one PBM noted
                               that not all utilization management services result in savings. For
                               example, they noted that improving care with medication therapy
                               management programs may increase drug costs through increased
                               utilization. Additionally, representatives of one Part D plan sponsor noted
                               the savings from utilization management services in commercial plans
                               may be greater than in Part D because the use of manufacturers’ copay
                               coupons are prohibited in federal health care programs, including Part
                               D. 20 While the coupons reduce or eliminate beneficiaries’ out-of-pocket
                               co-payments for certain brand-name drugs, thereby encouraging their
                               use, the coupons do not affect the amount that the plans pay for drugs.
                               Therefore, to the extent that beneficiaries in their commercial plans use
                               coupons, Part D plan sponsors have a greater incentive to employ
                               utilization management services in these plans to reduce the use of more
                               expensive brand-name drugs.

                               Representatives of Part D plan sponsors and PBMs we interviewed
                               differed with manufacturers and, in some cases, with each other on the
                               effects of utilization management services on various non-financial
                               aspects of drug utilization:

                               •     Beneficiary health. Representatives from all three manufacturers we
                                     interviewed stated that utilization management services negatively

                               19
                                 An industry-sponsored study provided to us by a PBM estimated that in 2018, PBMs
                               would save the Part D program $19 billion (or $35.15 per member per month) by
                               promoting the use of generic drugs. Actuarial Practice of Oliver Wyman, commissioned by
                               the Glover Park Group on behalf of the Coalition for Affordable Prescription Drugs,
                               Savings Generated By Pharmacy Benefit Managers in the Medicare Part D Program
                               (June 2017). The study did not specify how the financial savings from the estimated 5
                               percent increase in the generic dispensing rate due to PBMs was calculated.
                               20
                                   42 U.S.C. § 1320a-7b.




                               Page 68                                                     GAO-19-498 Medicare Part D
Appendix VI: Studies and Stakeholders’ Views
on Effects of Utilization Management Services




     affected beneficiary health by reducing their access to necessary
     medications. In contrast, seven of the 11 Part D plan sponsors and
     four of the five PBMs that discussed the effect of utilization
     management services on beneficiary health stated that utilization
     management services generally resulted in improved beneficiary
     health. 21 Representatives of certain PBMs and one Part D plan
     sponsor provided us examples of the ways utilization management
     services have improved their beneficiaries’ health, such as through
     opioid quantity limits. One Part D plan sponsor noted that point-of-sale
     utilization management services warn pharmacies of therapeutic
     duplications, toxicities across multiple prescriptions, or interactions of
     certain drugs with health conditions.
•    Medication access. Representatives from all three drug
     manufacturers noted that utilization management services impose
     limits on beneficiaries’ access to drugs, while seven of nine Part D
     plan sponsors and three of the four PBMs who discussed this stated
     utilization management services had no significant restrictions on
     beneficiaries’ access to necessary medications. Representatives from
     one plan sponsor noted there are appeals processes to ensure
     beneficiaries’ access is not adversely impacted by utilization
     management services.
•    Medication adherence. Representatives from all three manufacturers
     told us that utilization management services limit beneficiaries’
     adherence to their medications, such as by causing delays in therapy,
     while seven of eight Part D plan sponsors and all four PBMs who
     discussed this stated utilization management services had no adverse
     impact on beneficiaries’ adherence to their medications.
     Representatives from one plan sponsor and two PBMs stated that
     utilization management services may have a positive impact on
     adherence, such as by lowering copays through generic substitution.
•    Medicare protected classes and utilization management.
     Representatives from Part D plan sponsors, PBMs, and
     manufacturers differed in their views on the effect of Part D utilization
     management services restrictions on protected class drugs on
     beneficiary health. Representatives of two PBMs told us the effect
     was positive, as beneficiaries who use these drugs do not experience
     disruptions in therapy. Representatives of two other PBMs said there
     was no effect, and one said there was a negative effect—as plan

21
  Representatives we interviewed from four other Part D plan sponsors and one PBM said
utilization management services have no effect on beneficiaries’ health.




Page 69                                                    GAO-19-498 Medicare Part D
Appendix VI: Studies and Stakeholders’ Views
on Effects of Utilization Management Services




    sponsors were required to cover certain less effective drugs.
    Representatives of one PBM said that, for example, patients in
    commercial health plans do not have any problems accessing
    protected class drugs that are subject to utilization management.
    These representatives noted that the Centers for Medicare &
    Medicaid Services provides for adequate access. However, one
    manufacturer told us that utilization management services for HIV
    drugs are rightly restricted in Part D, as these services may cause
    disruptions in therapy, which can lead to drug resistance and poorer
    health outcomes.

Representatives of five Part D plan sponsors said Medicare’s restrictions
on the use of utilization management services for protected class drugs
have had a negative impact on beneficiary health because, for example,
they limit plans’ ability to ensure that a prescribed drug is appropriate,
such as ensuring that a cancer drug is appropriate for a beneficiary’s
weight. Another plan sponsor representative told us the restrictions may
have a positive impact by reducing increases in medical costs, while
another plan sponsor said the restrictions have had no impact.




Page 70                                             GAO-19-498 Medicare Part D
Appendix VII: Bibliography of Peer Reviewed
              Appendix VII: Bibliography of Peer Reviewed
              Studies Used in GAO’s Literature Review



Studies Used in GAO’s Literature Review

              Abbass, I. M., E. O. Caplan, D. B. Ng, R. Kristy, C. R. Schermer, P. Bradt,
              J. M. Collins, W. Man, M. Chan, and B. T. Suehs. “Impact of Overactive
              Bladder Step Therapy Policies on Medication Utilization and Expenditures
              among Treated Medicare Members.” Journal of Managed Care &
              Specialty Pharmacy, vol. 23, no. 1 (2017): 27-37.

              Agarwal, A., R. A. Freedman, F. Goicuria, C. Rhinehart, K. Murphy, E.
              Kelly, E. Mullaney, M. St. Amand, P. Nguyen, and N. U. Lin. “Prior
              Authorization for Medications in a Breast Oncology Practice: Navigation
              of a Complex Process.” Journal of Oncology Practice, vol. 13, no. 4
              (2017): e273-e282.

              Almodovar, A. S., D. R. Axon, A. M. Coleman, T. Warholak, and M. C.
              Nahata. “The Effect of Plan Type and Comprehensive Medication
              Reviews on High-Risk Medication Use.” Journal of Managed Care &
              Specialty Pharmacy, vol. 24, no. 5 (2018): 416-422.

              Alston, G., and C. Hanrahan. “Can a Pharmacist Reduce Annual Costs
              for Medicare Part D Enrollees?” The Consultant Pharmacist, vol. 26, no. 3
              (2011): 182-189.

              Ben-Joseph, R., C.C. Chen, A. P. De, R.L. Wade, and D. Shah.
              “Consequences of Patient Access Restrictions to Branded Oxycodone
              Hydrochloride Extended-Release Tablets on Healthcare Utilization and
              Costs in US Health Plans.” Journal of Medical Economics, vol. 17, no. 10
              (2014): 708-718.

              Bergeson, J. G., K. Worley, A. Louder, M. Ward, and J. Graham.
              “Retrospective Database Analysis of the Impact of Prior Authorization for
              Type 2 Diabetes Medications on Health Care Costs in a Medicare
              Advantage Prescription Drug Plan Population.” Journal of Managed Care
              Pharmacy, vol. 19, no. 5 (2013): 374-384.

              Bloudek, L. M., D. Makenbaeva, and M. Eaddy. “Anticipated Impact of
              Generic Imatinib Market Entry on the Costs of Tyrosine Kinase Inhibitors.”
              American Health & Drug Benefits, vol. 8, no. 9 (2015): 472-480.

              Branham, A., J. Moose, and S. Ferrari. “Retrospective Analysis of
              Medication Adherence and Cost Following Medication Therapy
              Management.” Innovation in Pharmacy, vol. 1, no. 1 (2010): 1-8.




              Page 71                                             GAO-19-498 Medicare Part D
Appendix VII: Bibliography of Peer Reviewed
Studies Used in GAO’s Literature Review




Branham, A. R., A. J. Katz, J. S. Moose, S. P. Ferreri, J. F. Farley, and M.
W. Marciniak. “Retrospective Analysis of Estimated Cost Avoidance
Following Pharmacist-Provided Medication Therapy Management
Services.” Journal of Pharmacy Practice, vol. 26, no. 4 (2013): 420-427.

Buhl, A., J. Augustine, A. M. Taylor, R. Martin, and T. L. Warholak.
“Positive Medication Changes Resulting from Comprehensive and
Noncomprehensive Medication Reviews in a Medicare Part D
Population.” Journal of Managed Care & Specialty Pharmacy, vol. 23, no.
3 (2017): 388-394.

Caffiero, N., T. Delate, M. D. Ehizuelen, and K. Vogel. “Effectiveness of a
Clinical Pharmacist Medication Therapy Management Program in
Discontinuation of Drugs to Avoid in the Elderly.” Journal of Managed
Care & Specialty Pharmacy, vol. 23, no. 5 (2017): 525-531.

Caplan, E. O., M. C. Guy, J. Chang, and K. Boesen. “Telephone-Based
Cardiovascular Medication Therapy Management in Medicare Part D
Enrollees with Diabetes.” The American Journal of Pharmacy Benefits,
vol. 9, no. 2 (2017): 47-54.

Chinthammit, C., E. P. Armstrong, K. Boesen, R. Martin, A. M. Taylor,
and T. Warholak. “Cost-Effectiveness of Comprehensive Medication
Reviews Versus Noncomprehensive Medication Review Interventions and
Subsequent Successful Medication Changes in a Medicare Part D
Population.” Journal of Managed Care & Specialty Pharmacy, vol. 21, no.
5 (2015): 381-389.

Chisholm-Burns, M. A., C. A. Spivey, E. A. Tolley, and E. K. Kaplan.
“Medication Therapy Management and Adherence among US Renal
Transplant Recipients.” Patient Preference and Adherence, vol.10 (2016):
703-709.

Corsi, K., V. Lemay, K. K. Orr, and L. Cohen. “Pharmacist Medication
Therapy Management in Home Health Care: Investigation of a
Sustainable Practice Model.” Journal of the American Pharmacists
Association, vol. 58, no. 4S (2018): S64-S68.

Dai, R. and J. Robst. “The Relationship between Plan Characteristics and
Medicare Prescription Drug Plan Bids.” Applied Economics Letters, vol.
19, no. 1 (2012): 99-104.




Page 72                                             GAO-19-498 Medicare Part D
Appendix VII: Bibliography of Peer Reviewed
Studies Used in GAO’s Literature Review




De Lott, L. B., J. F. Burke, K. A. Kerber, L. E. Skolarus, and B. C.
Callaghan. “Medicare Part D Payments for Neurologist-Prescribed
Drugs.” Neurology, vol. 86, no. 16 (2016): 1491-1498.

Dodson, S. E., J. F. Ruisinger, P. A. Howard, S. E. Hare, and B. J.
Barnes. “Community Pharmacy-Based Medication Therapy Management
Services: Financial Impact for Patients.” Pharmacy Practice, vol. 10, no. 3
(2012): 119-124.

Duru, O. K., S. L. Ettner, N. Turk, C. M. Mangione, A. F. Brown, J. Fu, L.
Simien, and C.W. Tseng. “Potential Savings Associated with Drug
Substitution in Medicare Part D: The Translating Research Into Action for
Diabetes (TRIAD) Study.” Journal of General Internal Medicine, vol. 29,
no. 1 (2014): 230-236.

Egilman, A. C., J. D. Wallach, J. S. Ross, and S. S. Dhruva. “Medicare
Spending and Potential Savings on Brand-Name Drugs with Available
Generic Substitutes Excluded by 2 Large Pharmacy Benefit Managers,
2012 through 2015.” JAMA Internal Medicine, vol. 178, no. 4 (2018): 567-
569.

Fox, D., L. D. Ried, G. E. Klein, W. Myers, and K. Foli. “A Medication
Therapy Management Program’s Impact on Low-Density Lipoprotein
Cholesterol Goal Attainment in Medicare Part D Patients with Diabetes.”
Journal of the American Pharmacists Association, vol. 49, no. 2 (2009):
192-199.

Gellad, W. F., J. M. Donohue, X. Zhao, M. K. Mor, C. T. Thorpe, J. Smith,
C. B. Good, M. J. Fine, and N. E. Morden. “Brand-Name Prescription
Drug Use among Veterans Affairs and Medicare Part D Patients with
Diabetes: A National Cohort Comparison.” Annals of Internal Medicine,
vol. 159, no. 2 (2013): 105-114.

Gernant, S. A., M. E. Snyder, H. Jaynes, J. M. Sutherland, and A. J.
Zillich. “The Effectiveness of Pharmacist-Provided Telephonic Medication
Therapy Management on Emergency Department Utilization in Home
Health Patients.” Journal of Pharmacy Technology, vol. 32, no. 5 (2016):
179-184.

Gold, J. A., B. French, and L. C. Vermeulen. “Reduction of Use of
Potentially Inappropriate Medications in the Elderly.” Wisconsin Medical
Journal, vol. 107, no. 4 (2008): 213-214.



Page 73                                             GAO-19-498 Medicare Part D
Appendix VII: Bibliography of Peer Reviewed
Studies Used in GAO’s Literature Review




Hoadley, J. F., K. Merrell, E. Hargrave, and L. Summer. “In Medicare Part
D Plans, Low or Zero Copays and Other Features to Encourage the Use
of Generic Statins Work, Could Save Billions.” Health Affairs, vol. 31, no.
10 (2012): 2266-75.

Hui, R. L., B. D. Yamada, M. M. Spence, E. W. Jeong, and J. Chan.
“Impact of a Medicare MTM Program: Evaluating Clinical and Economic
Outcomes.” The American Journal of Managed Care, vol. 20, no. 2
(2014): e43-e51.

Huskamp, H. A., D. G. Stevenson, A. J. O’Malley, S. B. Dusetzina, S. L.
Mitchell, B. J. Zarowitz, M. E. Chernew, and J. P. Newhouse. “Medicare
Part D Plan Generosity and Medication Use among Dual-Eligible Nursing
Home Residents.” Medical Care, vol. 51, no. 10 (2013): 894-900.

Moczygemba, L. R., J. C. Barner, J. C. Brannier, and E. R. Gabrillo.
“Outcomes of a Medicare Part D Telephone Medication Therapy
Management Program.” Journal of the American Pharmacists
Association, vol. 52, no. 6 (2012): e144-e152.

Moczygemba, L. R., J. C. Barner, K. A. Lawson, C. M. Brown, E. R.
Gabrillo, P. Godley, and M. Johnsrud. “Impact of Telephone Medication
Therapy Management on Medication and Health-Related Problems,
Medication Adherence, and Medicare Part D Drug Costs: A 6-Month
Follow Up.” The American Journal of Geriatric Pharmacotherapy, vol. 9,
no. 5 (2011): 328-338.

Moore, J. M., D. Shartle, L. Faudskar, O.S. Matlin, and T.A. Brennan.
Impact of a Patient-Centered Pharmacy Program and Intervention in a
High-Risk Group. Journal of Managed Care Pharmacy, vol. 19, no. 3
(2013): 228-236.

Newman-Casey, P. A., M. A. Woodward, L. M. Niziol, P. P. Lee, and L. B.
De Lott. “Brand Medications and Medicare Part D: How Eye Care
Providers’ Prescribing Patterns Influence Costs.” Ophthalmology, vol.
125, no. 3 (2018): 332-339.

Null, K. D., K. Moll, A. Sadosky, M. K. Pasquale, J. C. Cappelleri, and B.
Parsons. “Trends Associated with Implementing and Lifting a Pregabalin
Step Therapy Policy.” American Journal of Pharmacy Benefits, vol. 8, no.
2 (2016): e17-e24.




Page 74                                             GAO-19-498 Medicare Part D
Appendix VII: Bibliography of Peer Reviewed
Studies Used in GAO’s Literature Review




Olvey, E. L., M. C. Guy, J. Chang, and G. H. Skrepnek. “Cost-
Effectiveness of Medication Therapy Management in Part D Diabetic
Enrollees.” American Journal of Pharmacy Benefits, vol. 6, no. 5 (2014):
e147-e156.

Pai, A.B., A. Boyd, J. Depczynski, I.M. Chavez, N. Khan, and H. Manley.
Reduced Drug Use and Hospitalization Rates in Patients Undergoing
Hemodialysis Who Received Pharmaceutical Care: A 2-Year,
Randomized, Controlled Study. Pharmacotherapy, vol. 29, no. 12 (2009):
1433-1440.

Patel, R. A., M. P. Walberg, E. Tong, F. Tan, A. E. Rummel, J. A. Woelfel,
S. M. Carr-Lopez, and S. M. Galal. “Cost Variability of Suggested Generic
Treatment Alternatives under the Medicare Part D Benefit.” Journal of
Managed Care Pharmacy, vol. 20, no. 3 (2014): 283-290.

Pindolia, V. K., L. Stebelsky, T. M. Romain, L. Luoma, S. N. Nowak, and
F. Gillanders. “Mitigation of Medication Mishaps via Medication Therapy
Management.” The Annals of Pharmacotherapy, vol. 43, no. 4 (2009):
611-620.

Shen, X., B. C. Stuart, C. A. Powers, S. E. Tom, L. S. Magder, and E. M.
Perfetto. “Impact of Formulary Restrictions on Medication Use and
Costs.” The American Journal of Managed Care, vol. 23, no. 8 (2017):
e265-e274.

Starner, C. I., S. A. Norman, R. G. Reynolds, and P. P. Gleason. “Effect
of a Retrospective Drug Utilization Review on Potentially Inappropriate
Prescribing in the Elderly.” The American Journal of Geriatric
Pharmacotherapy, vol. 7, no. 1 (2009): 11-19.

Stebbins, M.R., D.J. Kaufman, and H.L. Lipton. “The PRICE Clinic for
Low-Income Elderly: A Managed Care Model for Implementing
Pharmacist-Directed Services.” Journal of Managed Care Pharmacy, vol.
11, no. 4 (2005): 333–341.

Steele, K. M., J. F. Ruisinger, J. Bates, E. S. Prohaska, B. L. Melton, and
S. Hipp. “Home-Based Comprehensive Medication Reviews:
Pharmacist’s Impact on Drug Therapy Problems in Geriatric Patients.”
The Consultant Pharmacist, vol. 31, no. 10 (2016): 598-605.

Stockl, K. M., D. Tjioe, S. Gong, J. Stroup, A. S. M. Harada, and H. C.
Lew. “Effect of an Intervention to Increase Statin Use in Medicare


Page 75                                             GAO-19-498 Medicare Part D
Appendix VII: Bibliography of Peer Reviewed
Studies Used in GAO’s Literature Review




Members Who Qualified for a Medication Therapy Management
Program.” Journal of Managed Care Pharmacy, vol. 14, no. 6 (2008):
532-540.

Suehs, B. T., A. Louder, M. Udall, J. C. Cappelleri, A. V. Joshi, and N. C.
Patel. 2014. “Impact of a Pregabalin Step Therapy Policy among
Medicare Advantage Beneficiaries.” Pain Practice, vol. 14 (5): 419-426.

Tang, Y., W. F. Gellad, A. Men, and J. M. Donohue. “Impact of Medicare
Part D Plan Features on Use of Generic Drugs.” Medical Care, vol. 52,
no. 6 (2014): 541-548.

Thatcher, E. E., E. M. Vanwert, and S. R. Erickson. “Potential Impact of
Pharmacist Interventions to Reduce Cost for Medicare Part D
Beneficiaries.” Journal of Pharmacy Practice, vol. 26, no. 3 (2013): 248-
252.

Touchette, D. R., A. L. Masica, R. J. Dolor, G. T. Schumock, Y. K. Choi,
Y. Kim, and S. R. Smith. “Safety-Focused Medication Therapy
Management: A Randomized Controlled Trial.” Journal of the American
Pharmacists Association, vol. 52, no. 5 (2012): 603-612.

Ward, M. A., and Y. Xu. “Pharmacist-Provided Telephonic Medication
Therapy Management in an MAPD Plan.” The American Journal of
Managed Care, vol. 17, no. 10 (2011): e399-e409.

Welch, E. K., T. Delate, E. A. Chester, and T. Stubbings. “Assessment of
the Impact of Medication Therapy Management Delivered to Home-Based
Medicare Beneficiaries.” The Annals of Pharmacotherapy, vol. 43, no. 4
(2009): 603-610.

West, J. C., J. E. Wilk, I. L. Muszynski, D. S. Rae, M. Rubio-Stipec, C. L.
Alter, W. E. Narrow, and D. A. Regier. “Medication Access and Continuity:
The Experiences of Dual-Eligible Psychiatric Patients during the First 4
Months of the Medicare Prescription Drug Benefit.” The American Journal
of Psychiatry, vol. 164, no. 5 (2007): 789-796.

West, J. C., J. E. Wilk, D. S. Rae, I. L. Muszynski, M. Rubio-Stipec, C. L.
Alter, K. E. Sanders, S. Crystal, and D. A. Regier. “First-Year Medicare
Part D Prescription Drug Benefits: Medication Access and Continuity
among Dual Eligible Psychiatric Patients.” The Journal of Clinical
Psychiatry, vol. 71, no. 4 (2010): 400-410.



Page 76                                              GAO-19-498 Medicare Part D
Appendix VII: Bibliography of Peer Reviewed
Studies Used in GAO’s Literature Review




Winston, S., and Y.S. Lin. “Impact on Drug Cost and Use of Medicare
Part D of Medication Therapy Management Services Delivered in 2007.”
Journal of the American Pharmacists Association, vol. 49, no. 6 (2009):
813-820.

Woelfel, J. A., S. M. Carr-Lopez, M. D. Santos, A. Bui, R. A. Patel, M. P.
Walberg, and S. M. Galal. “Assessing Medicare Beneficiaries’
Willingness-to-Pay for Medication Therapy Management Services.” The
Consultant Pharmacist, vol. 29, no. 2 (2014): 104-109.

Zillich, A. J., M. E. Snyder, C. K. Frail, J. L. Lewis, D. Deshotels, P.
Dunham, H. A. Jaynes, and J. M. Sutherland. “A Randomized, Controlled
Pragmatic Trial of Telephonic Medication Therapy Management to
Reduce Hospitalization in Home Health Patients.” Health Services
Research, vol. 49, no. 5 (2014): 1537-1554.




Page 77                                             GAO-19-498 Medicare Part D
Appendix VIII: GAO Contact and Staff
                   Appendix VIII: GAO Contact and Staff
                   Acknowledgments



Acknowledgments


                   John E. Dicken, (202) 512-7114 or dickenj@gao.gov
GAO Contact
                   In addition to the contact named above, Robert Copeland, Assistant
Staff              Director; William A. Crafton, Analyst-in-Charge; Britt Carlson, Kaitlin
Acknowledgements   Dunn, Andrew Emmons, Michael Rose, and Dan Ries made key
                   contributions to this report. Also contributing were George Bogart, Yesook
                   Merrill, Laurie Pachter, and Vikki Porter.




(102043)
                   Page 78                                            GAO-19-498 Medicare Part D
                         The Government Accountability Office, the audit, evaluation, and investigative
GAO’s Mission            arm of Congress, exists to support Congress in meeting its constitutional
                         responsibilities and to help improve the performance and accountability of the
                         federal government for the American people. GAO examines the use of public
                         funds; evaluates federal programs and policies; and provides analyses,
                         recommendations, and other assistance to help Congress make informed
                         oversight, policy, and funding decisions. GAO’s commitment to good government
                         is reflected in its core values of accountability, integrity, and reliability.

                         The fastest and easiest way to obtain copies of GAO documents at no cost is
Obtaining Copies of      through GAO’s website (https://www.gao.gov). Each weekday afternoon, GAO
GAO Reports and          posts on its website newly released reports, testimony, and correspondence. To
                         have GAO e-mail you a list of newly posted products, go to https://www.gao.gov
Testimony                and select “E-mail Updates.”

Order by Phone           The price of each GAO publication reflects GAO’s actual cost of production and
                         distribution and depends on the number of pages in the publication and whether
                         the publication is printed in color or black and white. Pricing and ordering
                         information is posted on GAO’s website, https://www.gao.gov/ordering.htm.
                         Place orders by calling (202) 512-6000, toll free (866) 801-7077, or
                         TDD (202) 512-2537.
                         Orders may be paid for using American Express, Discover Card, MasterCard,
                         Visa, check, or money order. Call for additional information.

                         Connect with GAO on Facebook, Flickr, Twitter, and YouTube.
Connect with GAO         Subscribe to our RSS Feeds or E-mail Updates. Listen to our Podcasts.
                         Visit GAO on the web at https://www.gao.gov.

                         Contact FraudNet:
To Report Fraud,
                         Website: https://www.gao.gov/fraudnet/fraudnet.htm
Waste, and Abuse in
                         Automated answering system: (800) 424-5454 or (202) 512-7700
Federal Programs
                         Orice Williams Brown, Managing Director, WilliamsO@gao.gov, (202) 512-4400,
Congressional            U.S. Government Accountability Office, 441 G Street NW, Room 7125,
Relations                Washington, DC 20548

                         Chuck Young, Managing Director, youngc1@gao.gov, (202) 512-4800
Public Affairs           U.S. Government Accountability Office, 441 G Street NW, Room 7149
                         Washington, DC 20548

                         James-Christian Blockwood, Managing Director, spel@gao.gov, (202) 512-4707
Strategic Planning and   U.S. Government Accountability Office, 441 G Street NW, Room 7814,
External Liaison         Washington, DC 20548




                            Please Print on Recycled Paper.