MedPAC, Hospital inpatient and outpatient services (March 2023 Report to the Congress, ch. 3) (MedPAC March 2023 ch. 3, 2023-03)
Summary
Chapter 3, "Hospital inpatient and outpatient services," of MedPAC's Report to the Congress: Medicare Payment Policy, March 2023. Recommendation 3-1 says that for fiscal year 2024 the Congress should update the 2023 Medicare base payment rates for general acute care hospitals by the amount specified in current law plus 1 percent. Recommendation 3-2 would begin a transition to redistribute disproportionate share hospital and uncompensated care payments through the Medicare Safety-Net Index (MSNI) and add $2 billion to the MSNI pool. Commissioner votes on each are printed as YES 17, NO 0, NOT VOTING 0, ABSENT 0. The chapter summary says the fee-for-service Medicare program and beneficiaries paid general acute care hospitals $182.5 billion for inpatient and outpatient services in 2021.
Summary drafted by a model from the document's text below and checked by script against that text before publication. It is a navigation aid, not a reading of what the document proves. Where AI is used
Cite this document
“MedPAC, Hospital inpatient and outpatient services (March 2023 Report to the Congress, ch. 3) (MedPAC March 2023 ch. 3, 2023-03).”
Original at www.medpac.gov. Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/medpac-hospital-inpatient-and-outpatient-services-march-2023-report-to-f1677a0402/
, SHA-256 83db294897526fce429cbf2c647f3218afbb4c991746949e09749202809b21d0.Full text
C H A P T E R
3
Hospital inpatient and
outpatient services
R E C O M M E N D A T I O N S
3-1 For fiscal year 2024, the Congress should update the 2023 Medicare base payment
rates for general acute care hospitals by the amount specified in current law plus
1 percent.
COMMISSIONER VOTES: YES 17 • NO 0 • NOT VOTING 0 • ABSENT 0
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3-2 In fiscal year 2024, the Congress should:
• begin a transition to redistribute disproportionate share hospital and
uncompensated care payments through the Medicare Safety-Net Index (MSNI);
• add $2 billion to the MSNI pool;
• scale fee-for-service MSNI payments in proportion to each hospital’s MSNI
and distribute the funds through a percentage add-on to payments under the
inpatient and outpatient prospective payment systems; and
• pay commensurate MSNI amounts for services furnished to Medicare
Advantage (MA) enrollees directly to hospitals and exclude them from MA
benchmarks.
COMMISSIONER VOTES: YES 17 • NO 0 • NOT VOTING 0 • ABSENT 0
Hospital inpatient and
outpatient services
C H A P T E R
3
Chapter summary In this chapter
General acute care hospitals (ACHs) primarily provide inpatient care
• Are Medicare payments
and various outpatient services. To pay these hospitals for their facility adequate in 2023?
costs, fee-for-service (FFS) Medicare generally sets prospective payment
rates under the inpatient prospective payment systems (IPPS) and the • How should Medicare
payments change in 2024?
outpatient prospective payment system (OPPS). In 2021, the FFS Medicare
program and its beneficiaries paid general acute care hospitals $182.5 • Supporting Medicare
billion for inpatient and outpatient services under the IPPS and OPPS, safety-net hospitals
including $8.3 billion in uncompensated care payments made under the
• Appendix: Supplemental
IPPS.
information on the
Medicare Safety-Net Index
Assessment of payment adequacy
In 2021, most indicators of hospital payment adequacy remained positive
or improved. However, indicators continued to vary substantially across
hospitals, and some indicators remained below prepandemic levels. In
2022, input cost increases for hospitals were higher and more volatile
than they have been in recent years.
Beneficiaries’ access to care—In 2021 and 2022, the number of general
ACHs that closed was the same as the number that opened, hospitals
continued to have excess capacity in aggregate, and those with excess
capacity continued to have a financial incentive to serve FFS Medicare
Report to the Congress: Medicare Payment Policy | March 2023 55
beneficiaries. However, some hospitals faced occupancy and staffing
constraints at times. In 2021, IPPS hospitals’ marginal profit on IPPS and OPPS
services (a measure of whether providers have a financial incentive to expand
the number of Medicare beneficiaries they serve) was about 8 percent, which is
similar to prepandemic levels.
Quality of care—In 2021, FFS beneficiaries’ risk-adjusted hospital readmission
rate improved relative to 2019. However, the risk-adjusted hospital mortality
rate remained higher than in 2019, and most patient experience measures
declined.
Providers’ access to capital—Hospitals’ access to capital strengthened in 2021,
as IPPS hospitals’ all-payer operating margin reached a record high of 8.7
percent. However, there was substantial variation in margins across hospitals.
Preliminary data indicate that 2022 all-payer operating margins declined
relative to 2021 and were mixed relative to prepandemic levels, but most
hospitals continued to have strong access to bond markets.
Medicare payments and providers’ costs—In 2021, Medicare’s payments to
hospitals continued to be below hospitals’ costs in aggregate but near costs
among relatively efficient hospitals and higher than in 2020. IPPS hospitals’
Medicare margin increased in 2021 to –6.2 percent when including a share of
federal relief funds (–8.3 percent exclusive of these funds), and the median
Medicare margin for relatively efficient hospitals increased to 1 percent
(near break-even exclusive of federal relief funds). However, we project that
hospitals’ Medicare margins in 2023 will be lower than in 2021, driven in part
by growth in hospitals’ input costs, which exceeded the forecasts CMS used
to set Medicare payment rate updates, and in part by the expected expiration
of federal relief funds and temporary Medicare payment increases related to
the public health emergency. These federal relief funds and Medicare payment
increases exceeded hospitals’ additional costs related to coronavirus disease
2019 (COVID-19). We anticipate that reductions in net revenue will be partially
offset by other factors, including (1) reductions in hospitals’ costs related to
COVID-19, as cases decline and hospitals become better at managing the
disease; and (2) the statutory 0.5 percent increase to inpatient operating
payments to remove prior temporary reductions for past documentation
and coding changes. We estimate that IPPS hospitals’ Medicare margin will
decrease in 2023 to about –10 percent (similar to the level in 2017) and that
the median Medicare margin for relatively efficient hospitals will decrease to
modestly below break-even—similar to prepandemic levels.
56 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
How should Medicare payments change in 2024?
The current-law updates to payment rates for 2024 will not be finalized until
summer 2023, but CMS’s third-quarter 2022 forecasts would result in the IPPS
operating base payment rate and OPPS base payment rate increasing by 2.9
percent and the IPPS capital base payment rate increasing by 2.4 percent.
The Commission anticipates that a 2024 update to hospital payment rates
of current law plus 1 percent would generally be adequate to maintain FFS
beneficiaries’ access to hospital inpatient and outpatient care and keep IPPS
and OPPS payment rates close to the cost of delivering high-quality care
efficiently. The Commission’s payment update recommendation for 2024
reflects the most recent inflation and other data from 2021, preliminary data
from 2022, and projections for 2023. If current projections of input inflation and
hospital costs turn out to be inaccurate, these discrepancies will be accounted
for in our assessment of payment adequacy in our next recommendation cycle.
Supporting Medicare safety-net hospitals
The recommended update to IPPS and OPPS payment rates of current law plus 1
percent may not be sufficient to ensure the financial viability of some Medicare
safety-net hospitals with a poor payer mix. As the Medicare program strives
to ensure access to care for all beneficiaries and adequately pay providers for
that access, additional Medicare payments to Medicare safety-net providers are
warranted. Medicare already provides substantial safety-net funding to hospitals,
but there are several problems with the way Medicare distributes these funds,
including omitting a hospital’s Medicare share from its funding formulas in
favor of subsidizing Medicaid payments, making supplemental payments only
for inpatient services, and having an uncompensated care payment formula that
favors hospitals with few FFS Medicare patients. The Commission’s view is that
Medicare safety-net payments should be used primarily to support Medicare
safety-net hospitals—those that provide care to large shares of low-income
Medicare beneficiaries. We note that this definition of “safety-net hospital” is
Medicare-centric by design; safety-net definitions used by Medicaid and other
payers would likely differ.
The Commission recommends redistributing the current Medicare safety-
net payments (disproportionate share hospital and uncompensated care
payments) using the Commission-developed Medicare Safety-Net Index
(MSNI) for hospitals. Implementation of this index would better target scarce
Medicare resources to support hospitals that are key sources of care for low-
Report to the Congress: Medicare Payment Policy | March 2023 57
income Medicare beneficiaries and may be at risk of closure. In addition to the
redistribution, the Commission recommends adding $2 billion to this MSNI
pool of funds to help maintain the financial viability of Medicare safety-net
hospitals. The FFS portion of the MSNI pool of funds should be distributed
to hospitals as add-on payments to Medicare’s IPPS and OPPS payments,
with commensurate add-on amounts made to hospitals treating Medicare
Advantage enrollees.
While most hospitals will see increases in Medicare revenue due to the $2
billion in additional Medicare safety-net spending, there are some hospitals
that will see reductions. Material reductions in Medicare revenue could
occur for hospitals that currently receive high Medicare uncompensated care
payments but serve relatively few FFS Medicare patients. In light of these
effects, the Congress could phase in the policy for all hospitals over a set period
of time (i.e., transition to the MSNI policy over three to five years). Alternatively,
a transition could be managed through a stop-loss policy so that no hospital
would experience changes (positive or negative) in Medicare payments of
more than 5 percent in any one year due to the transition to the MSNI. Both
approaches would also allow time for the hospitals facing the most substantial
revenue reductions to try to augment revenues from existing sources and
request additional financial support from state and local governments, as
warranted. To the extent that these hospitals have high cost structures, a
transition also would allow time to improve efficiencies. ■
58 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
outpatient services under the IPPS and OPPS, including
Background $8.3 billion in uncompensated care payments and $16.4
billion for separately payable drugs (Table 3-1).3,4
General acute care hospitals (ACHs) primarily provide
inpatient care and various outpatient services. To pay Medicare uses different payment methodologies to
these hospitals for their facility costs, fee-for-service reimburse certain other general ACHs for their facility
(FFS) Medicare generally sets prospective payment costs of providing inpatient and outpatient services
rates under the inpatient prospective payment systems to FFS beneficiaries. Most notably, Medicare has
(IPPS) and outpatient prospective payment system designated about 1,350 small rural hospitals as critical
(OPPS).1 (Clinicians who provide services at hospitals access hospitals (CAHs) and pays these hospitals based
are paid separately under the physician fee schedule; on their costs.5 There are also about 50 general ACHs
Medicare also pays separately for certain hospital in Maryland that Medicare pays based on an all-payer
units and costs.2) In setting these prospective rates global budget. These payment methodologies are
per inpatient stay or primary outpatient service, beyond the scope of this chapter.
CMS adjusts IPPS and OPPS national base payment
rates for factors outside of hospitals’ control, such as The IPPS and OPPS payment rates affect more than
regional wage rates and patient characteristics. Both FFS Medicare payments for general ACHs. Within
the IPPS and OPPS also include separate payments the FFS Medicare program, the OPPS is used to pay
not tied to the base payment rates. The IPPS includes for outpatient services at certain specialty hospitals
uncompensated care payments to help support and other facilities.6 But more importantly, most
hospitals’ costs of treating the uninsured. The OPPS Medicare Advantage plans pay IPPS hospitals using
sets payments for drugs that exceed a cost threshold rates benchmarked to FFS Medicare rates (Berenson
based on the manufacturer’s average sales price. In et al. 2015, Maeda and Nelson 2017). In addition, other
2021, the FFS Medicare program and its beneficiaries payers—such as the Department of Veterans Affairs,
paid general ACHs $182.5 billion for inpatient and certain state employee health plans, and some state
TABLE
Medicare pays for most acute care hospital services
3– 1 provided to FFS beneficiaries under the IPPS and OPPS, 2021
Number of FFS
Number of inpatient stays or
Number of FFS users outpatient services Payments
Medicare payment system hospitals (in millions) (in millions) (in billions)
IPPS—Inpatient stays 3,170 4.6 7.1 $107.9
IPPS—Uncompensated care 2,640 N/A N/A 8.3
OPPS—Outpatient services 3,370 17.1 135.7 49.9
OPPS—Separately payable drugs 16.4
Total 182.5
Note: FFS (fee-for-service), IPPS (inpatient prospective payment systems), OPPS (outpatient prospective payment system), N/A (not applicable). The
number of general acute care hospitals that provided OPPS services is higher than the number that provided IPPS services primarily because
about 200 facilities gained hospital provider numbers during the public health emergency but did not provide any inpatient services to FFS
beneficiaries. Number of hospitals rounded to the nearest 10. “OPPS—Separately payable drugs” includes drugs, devices, blood products, and
brachytherapy sources. Payments include applicable beneficiary cost-sharing responsibilities. “Year” refers to fiscal year for inpatient services
and calendar year for outpatient services.
Source: MedPAC analysis of Medicare Provider Analysis and Review data, IPPS final rule, and outpatient claims.
Report to the Congress: Medicare Payment Policy | March 2023 59
public option plans—also set hospitals’ rates based on The number of general ACHs that closed was the
FFS Medicare payments (Government Accountability same as the number that opened in both 2021 and
Office 2013, Schramm and Aters 2021, Scott 2021). 2022, hospitals continued to have excess capacity in
Given the widespread use of FFS Medicare payment aggregate, and those with excess capacity continued
rates as a benchmark, any update to the Medicare base to have a financial incentive to serve FFS beneficiaries.7
payment amount will affect many other payers (White However, some hospitals faced occupancy and staffing
et al. 2013). constraints at times.
Hospital care also accelerated its shift from inpatient
to outpatient settings. In 2021, inpatient stays per FFS
Are Medicare payments adequate beneficiary declined, remaining below the prepandemic
in 2023? trend, while the number of hospital outpatient services
per FFS beneficiary increased, reaching prepandemic
In 2021—the most recent year of data for most of our levels.
measures—most hospital payment adequacy indicators
remained positive or improved, despite the continued Supply of hospitals has been steady
coronavirus pandemic. In particular, the number of In both fiscal years 2021 and 2022, the number of
general ACHs that closed was the same as the number general ACHs that closed was the same as the number
that opened; IPPS hospitals’ all-payer operating that opened: 11 in 2021 and 16 in 2022.8 The number
margin increased to a record high of 8.7 percent; and of closures was substantially below the levels in 2019
the median Medicare margin for relatively efficient (46) and 2020 (25) and comparable with the number
hospitals increased to near break-even, exclusive of in 2017 and 2018. In contrast, from 2017 through 2022,
Medicare’s share of federal relief funds, and remained the number of hospital openings was steadier, ranging
at 1 percent when including these funds. from 8 to 18 openings.
However, hospital payment adequacy indicators Of the 16 hospitals that closed and the 16 that opened
continued to vary substantially across hospitals, and in fiscal year 2022, most shared several characteristics.9
some indicators remained below prepandemic levels. All were IPPS hospitals, most were in metropolitan
For example, some hospitals faced capacity and staffing areas (12 of 16 closures and 13 of 16 openings), and the
constraints at times. In addition, FFS beneficiaries’ risk- majority had 100 or fewer beds (9 of 16 closures and 11
adjusted hospital mortality rate remained higher than of 16 openings). In addition, almost all the closures (14
the rate in 2019. of 16) were within 25 miles of the next nearest hospital,
suggesting that most beneficiaries continued to have
In addition, in 2022, input cost increases for hospitals
access to inpatient and emergency services in their
were higher and more volatile than they have been
region, but some may have faced moderately longer
in recent years. Preliminary data from 2022 suggest
travel times.
that hospital margins were lower in 2022 than in 2021,
driven in part by higher-than-expected input costs. Medicare’s payment policies were not a main
contributor to the financial difficulties of the hospitals
For 2023, we project IPPS hospitals’ Medicare margin
that closed. Rather, many hospitals that closed in 2022
will decrease to about –10 percent (similar to the level
cited other financial reasons, such as failing to secure
in 2017), and the median Medicare margin for relatively
a buyer or low patient volume, as a driving factor for
efficient hospitals will decline to modestly below
the closure. Rural hospitals often face the greatest
break-even, similar to prepandemic levels.
challenges with declining admissions, in part due to
rural beneficiaries increasingly bypassing their local
Beneficiaries maintained good access to
hospitals to seek care at urban hospitals. However, as
hospital inpatient and outpatient services,
but some hospitals faced constraints at the Rural Emergency Hospital (REH) designation began
times on January 1, 2023, some rural hospitals in financial
distress may choose to convert to REHs rather than
FFS Medicare beneficiaries maintained good access
cease providing all services.10
to inpatient and outpatient services at general ACHs:
60 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
FIGURE
FIGU R Title here....
1-XX
3–1 In 2021, general acute care hospitals continued to have excess inpatient
capacity in aggregate, but some hospitals reached near capacity
100
80
Occupancy rate
60 65%
63% 64% 64%
62%
40
20
0
2017 2018 2019 2020 2021
Aggregate Middle half of hospitals Middle 90% of hospitals
Note: Data are for general acute care hospitals that had a cost report with a midpoint in the fiscal year and the report was complete as of our analysis.
Source: MedPAC analysis of hospital cost report data.
Note: Note and Source are in InDesign.
Hospitals had excess inpatient capacity in
Source: their capacity at times during 2022.11 At the other
aggregate and increased staff in 2021, but some extreme, a quarter of hospitals had an occupancy rate
hospitals faced capacity and staffing constraints of less than 30 percent, and 5 percent had occupancy
Notes about this graph:
at times rates below 15 percent, consistent with prior years.
General• Data
ACHs is in the to
continued atasheet. Make
have excess updates in the datasheet.
inpatient Some of these hospitals, which tended to be small
capacity in aggregate, with about 65 percent of all
• I deleted the years from the x-axis and put in myand own.rural, may not have sufficient inpatient volume
bed-days occupied during fiscal year 2021, slightly to maintain inpatient services, suggesting that they
higher•than
I had to manually
in prior draw3-1).
years (Figure tick This
marks and axis linescould
increased because they kept resetting when I changed any data.
be good candidates for the new Medicare REH
• Theresulted
occupancy dashedfromlinean
looked ok in
increase here, so I days
inpatient didn’t hand draw it.
esignation.
and a decrease in staffed inpatient beds.
• I can’t delete the legend, so I’ll just have to cropAnotherit out incomponent
InDesign. of hospitals’ capacity is their
However, inpatient
• Use direct capacity continued
selection tool to to vary
select items for modification. Otherwise
staffing level, which if you use
increased the in
black
slightly selection
2021 but wastool, they w
substantially across hospitals (Figure
default when you change the data. 3-1). For example, still below prepandemic trends. Prior to the pandemic,
in 2021, 5 percent of hospitals had occupancy rates hospital employment had been increasing by about 1
• Use paragraph styles (and object styles) to format.
of over 85 percent, which was slightly higher than percent a year; employment then declined by over 1
prior years. Moreover, as hospitals’ occupancy rates percent in fiscal year 2020 and recovered about half of
varied throughout the year, many of these hospitals this reduced employment in 2021.
likely neared or exceeded their capacity at times, and
preliminary data suggest that more hospitals exceeded Despite this modest increase in hospital employment,
some hospitals reported critical staffing shortages at
Report to the Congress: Medicare Payment Policy | March 2023 61
FIGURE
FIGU RE Title here....
31-XX
–2 Some hospitals reported critical staffing shortages
at times in fiscal year 2021 and, to a lesser extent, in 2022
25
Share of hospital days for which hospitals
20
15
reported a critical staffing shortage
10
5
0
Oct Dec Feb Apr Jun Aug Oct Dec Feb Apr Jun Aug
2020 2020 2021 2021 2021 2021 2021 2021 2022 2022 2022 2022
Note: Hospitals report critical staffing shortages to the Department of Health and Human Services based on their own individual facility needs and
staffing ratio policies. The use of temporary staff does not automatically count as having a staffing shortage. Data include all general acute care
hospitals and certain specialty hospitals and units.
Source: MedPAC analysis of healthdata.gov hospital capacity data.
Note: Note and Source are in InDesign.
Source:
times during fiscal year 2021 and, to a lesser extent, Hospitals with excess capacity continued to have
in 2022 (Figure 3-2). Throughout fiscal year 2021, a financial incentive to provide inpatient and
Notes
hospitals about
reported thisstaffing
a critical graph: shortage for outpatient services to FFS beneficiaries
over 10•percent
Data isofinall
thehospital days. Make
datasheet. The share of
updates In 2021, IPPS hospitals’ marginal profit on IPPS
in the datasheet.
hospital days with a critical staffing shortage was and OPPS services was about 8 percent—similar to
• I deleted the years from the x-axis
higher at the start and end of 2021, exceeding 15 and put in my own.
prepandemic levels. We calculate hospitals’ Medicare
percent• Iinhad
aggregate and over
to manually draw30 tick
percent
marks in some
and axis lines because
marginal theybykept
profit resetting
comparing when I IPPS
Medicare’s changed
and any data.
states. Anecdotal reports suggest that these staffing OPPS payments with the variable cost of treating
• The dashed line looked ok here, so I didn’t hand draw it.
shortages caused some hospitals to temporarily close an additional FFS Medicare patient. To make a
• I can’t delete
their emergency the legend,
departments so I’ll just
or intensive carehave
unitsto crop it out in InDesign.
conservative estimate of hospitals’ Medicare marginal
and to• Use
postpone or delay
direct certain
selection services
tool to select such as for modification.
items profit, we use a broad if you
Otherwise definition
use of
thevariable
black costs that tool, they w
selection
surgeries. To address these staffing
default when you change the data. shortages, many is consistent with our prior estimates of the share
hospitals have increased their use of travel nurses of costs that varied over a one-year period. We have
• Use paragraph styles (and object
(Adegbesan 2022). However, in fiscal year 2022,
styles) to format.
found that roughly 80 percent of costs are variable,
reported critical staffing shortages declined; hospitals including in 2021; to the extent that a higher share of
reported a critical staffing shortage for around 5 hospitals’ costs are fixed, the marginal profit would be
percent of all hospital days since March 2022. higher.
62 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
FIGU R
FIGURE
Title here....
3– 3
1-XX In 2021, FFS beneficiaries’ inpatient stays at general acute care
hospitals declined, remaining well below the prepandemic trend
12 300
Aggregate inpatient stays (in millions)
Stays per 1,000 FFS beneficiaries
10 250
252 249
9.7 9.5 241
8 9.2 200
212
208
7.9
7.4
6 150
–1.8%
–6.1%
4 100
2 50
0 0
2017 2018 2019 2020 2021 2017 2018 2019 2020 2021
Note: FFS (fee-for-service). Analysis includes FFS Medicare beneficiaries’ stays at hospitals paid under the inpatient prospective payment systems,
critical access hospitals, and acute care hospitals in Maryland and U.S. territories.
Source: MedPAC analysis of Medicare Provider Analysis and Review data and the Medicare Trustees report.
As we noted last year, the rapid response to the steady throughout 2021, at a level similar to the end of
Note: Note and Source are in InDesign.
coronavirus pandemic has demonstrated that—in fiscal year 2020 (data not shown).
response to lower volumes—many hospitals can
Source:
substantially lower their costs over a matter of months. In contrast to the decline in the number of inpatient
We expect thatabout
hospitals will graph:
have an even greater stays per FFS beneficiary, in 2021, FFS beneficiaries’
Notes this average length of stay increased by 6.1 percent to 5.5
ability to adjust costs to patient volume when they have
• Data is in the datasheet. Make
a longer period to adjust to environmental changes and updates in the days (Figure 3-4, p. 64, left panel). The increase in
datasheet.
the • I
resulting long-term changes the average length of stay was driven by a 7.3 percent
deleted the years frominthevolume
x-axisthat
andcan bein my own.
put
anticipated. decrease in stays of 2 to 3 days and a 9.9 percent
• I had to manually draw tick marks and axis linesincrease because they kept
in stays resetting
of longer when
than one I changed
week (Figure any data.
In 2021,• The
FFSdashed line looked
beneficiaries’ ok here,
inpatient staysso I didn’t hand3-4, p. 64,
draw it.right panel). The increase in length of
declined, though their average length of stay stay was even larger in stays of over two weeks (17.7
• I can’t delete the legend, so I’ll just have to crop it out in InDesign.
increased percent) and over two months (19.9 percent) (data not
• Use direct selection tool to
From 2020 to 2021, the number of inpatient stays by
select items for modification. Otherwise the
shown). Collectively, if you use in
decline theinpatient
black selection
stays pertool, they w
default when
FFS Medicare you change
beneficiaries the ACHs
at general data. declined by beneficiary and the increase in average length of stay
6.1 • Useto 7.4
percent paragraph
million staysstyles(Figure
(and 3-3,
object
leftstyles) resulted in the total number of days per beneficiary
panel).to format.
Controlling for the number of FFS beneficiaries, the rebounding to near the prepandemic trend.
number of inpatient stays declined by 1.8 percent, to
From 2020 to 2021, the combination of the accelerated
208 stays per 1,000 FFS beneficiaries (Figure 3-3, right
drop in inpatient stays per FFS beneficiary and the rise
panel).12 Inpatient stays per beneficiary were relatively
Report to the Congress: Medicare Payment Policy | March 2023 63
FIGU R
FIGURE
Title here....
3– 4
1-XX In 2021, FFS beneficiaries’ average length of stay at general
acute care hospitals increased, driven by increase in stays of over 1 week
8 120
Stays per 1,000 FFS Medicare beneficiaries,
100 2 or 3 days
Average length of stay
6 4 to 7 days –7.3%
80
5.5
5.1
4.9 4.9 4.9
4 60
(in days)
9.9%
by length of stay
6.1%
>1 week
40
2 1 day
20
0 0
2017 2018 2019 2020 2021 2017 2018 2019 2020 2021
Note: FFS (fee-for-service). Data include FFS Medicare beneficiaries’ stays at hospitals paid under the inpatient prospective payment systems, critical
access hospitals, and acute care hospitals in Maryland and U.S. territories.
Source: MedPAC analysis of Medicare Provider Analysis and Review data and the Medicare Trustees report.
in average length of stay was driven by the acceleration settings. In contrast, from 2020 to 2021, the
Note: Note and Source are in InDesign.
of two trends related to the shift of certain care from number of inpatient stays per beneficiary with
inpatient to outpatient settings:
Source: a resource weight of greater than 3 increased
by 4 percent. (In 2021, the most common FFS
• Accelerated decline in less resource-intensive inpatient stays with a weight of greater than 3
Notes stays
inpatient about and this graph:
increase in more resource- were stays for infectious diseases with operating
intensive
• Data inpatient stays. From Make
is in the datasheet. 2020 toupdates
2021, thein the datasheet.
room procedures and MCCs, septicemia or severe
number of inpatient stays per FFS beneficiary sepsis with mechanical ventilation for more
• I adeleted
with Medicare theseverity–diagnosis
years from the x-axis related and put in my own.
group than 96 hours, and percutaneous cardiovascular
(MS–DRG)
•I had toweight of lessdraw
manually than 1tick
declined
marksby 12.3axis lines because
and they with
procedures keptdrug-eluting
resetting when I changed
stents any data.
and MCCs.)
percent, about twice as fast as prepandemic
• The dashed line looked ok here, so I didn’t hand draw it.
trends (Figure 3-5, left panel). (The MS–DRG • Accelerated decline in short inpatient stays for
• I can’t
weight delete
reflects the estimate
CMS’s legend, of sothe
I’ll just have to crop it out in
relative InDesign.conditions and an increase in
musculoskeletal
average
• resource
Use direct intensitytool
selection of atotype of stay.
select In for modification.
items long stays for
Otherwise respiratory
if conditions.
you use the blackFrom 2020 tool, they w
selection
2021, the most common FFS Medicare
default when you change the data. inpatient to 2021, the number of inpatient stays per FFS
stays with a weight of less than 1 were those for beneficiary for musculoskeletal conditions, such
• Use paragraph styles (and object styles) to format. as joint replacements, declined 14.5 percent, about
gastrointestinal hemorrhage, esophagitis without
major complications or comorbidities (MCCs), four times faster than before the pandemic (Figure
and kidney and urinary tract infections without 3-5, right panel). The decline was over three times
MCCs.) These less resource-intensive conditions larger (–50.5 percent) among the most common
can increasingly be treated in hospital outpatient type of musculoskeletal stay—major hip or knee
64 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
FIGU R
FIGURE
Title here....
3– 5
1-XX Increasingly, fewer FFS beneficiaries’ inpatient stays are
for low resource-intensive and musculoskeletal conditions
140 60
Stays per 1,000 FFS Medicare beneficiaries, Stays per 1,000 FFS Medicare beneficiaries,
Circulatory
120 1 to 2
50
100
40
Low (< 1) Musculoskeletal 11.2%
80
Respiratory
–12.3% 30
by resource weight by diagnosis category
60
Infectious
20
2 to 3 4.0%
40 –14.5%
High (>3) 10
20
0 0
2017 2018 2019 2020 2021 2017 2018 2019 2020 2021
Note: FFS (fee-for-service). Data include FFS Medicare beneficiaries’ stays at hospitals paid under the inpatient prospective payment systems (IPPS),
critical access hospitals, and acute care hospitals in Maryland and U.S. territories. Resource weight refers to the diagnosis-related group weight
used in the IPPS. Right panel shows the four most common major diagnostic categories in 2021.
Source: MedPAC analysis of Medicare Provider Analysis and Review data and the Medicare Trustees report.
joint replacement without MCCs—which had an In 2021, FFS beneficiaries’ hospital outpatient
average
Note:
length of stay of two days. Increasingly,
Note and Source are in InDesign. services increased, with services per beneficiary
these procedures can be safely provided in
Source:
reaching prepandemic levels
outpatient settings. In contrast, the number of From 2020 to 2021, the number of FFS beneficiaries’
inpatient stays per beneficiary for respiratory hospital outpatient services at hospitals covered under
Notes about this graph:
conditions, including COVID-19, increased 11.2 the OPPS, CAHs, and Maryland hospitals increased by
• DataThe
percent. is in the datasheet.
increase was Make
nearly 14 timesupdates
largerin the 12.9
datasheet.
percent to 159 million (Figure 3-6, p. 66, left panel).
(156 percent) among the most common type
• I deleted the years from the x-axis and put in myControlling own. for the number of FFS beneficiaries, the
of respiratory stay, respiratory infections and number of hospital outpatient services climbed 18.1
• I had to manually
inflammations with MCC, draw tickhad
which marks and axis linespercent
an average because they
to 5.2 kept resetting
services when I(Figure
per beneficiary changed3-6,any
p. data.
length
• Theof stay
dashed of 7linedays.looked ok here, so I didn’t hand66,draw rightit.
panel). The volume of outpatient services was
lower in January 2021 and February 2021 relative to the
Some•ofI the can’t deleteinthe
increase legend,
average so of
length I’ll stay
just may
have to crop it out in InDesign.
rest of the calendar year. From March 2021 through
also •stemUse from staffing
direct constraints
selection tool toat select
skilleditems
nursing
for modification.
December Otherwise if you volume
2021, outpatient use thewas
black selection tool, they w
steady.
facilities that limited hospitals’
default when you change the data. ability to discharge
patients to post-acute care facilities. This increase in FFS Medicare beneficiaries’ outpatient
• Use paragraph styles (and object styles) to format. hospital services per beneficiary was primarily driven
Preliminary data for 2022 suggest that the number of by increases in:
inpatient stays per FFS beneficiary declined at a faster
rate than in 2021 and that growth in average length of • COVID-19-related services. In 2021, there were 9.7
stay slowed. million services, or 0.3 per beneficiary, for COVID-
19-related care, including vaccine administration,
Report to the Congress: Medicare Payment Policy | March 2023 65
FIGU R
FIGURE
Title here....
3– 6
1-XX In 2021, FFS beneficiaries’ hospital outpatient services and services
per beneficiary increased, the latter nearing prepandemic levels
250 8
Aggregate hospital outpatient services
200
Hospital outpatient services
6
174 174 175
150 159 5.2 5.2 5.3 5.2
141 4 4.4
(in millions)
per FFS beneficiary
100
12.9%
18.1%
2
50
0 0
2017 2018 2019 2020 2021 2017 2018 2019 2020 2021
Note: FFS (fee-for-service). Data include FFS Medicare beneficiaries’ outpatient services at hospitals paid under the outpatient prospective payment
system, critical access hospitals, and acute care hospitals in Maryland.
Source: MedPAC analysis of outpatient claims data and the Medicare Trustees report.
specimen collection, and chest X-rays. Since paid under the OPPS declined from 43 to 32 per 1,000
Note: Note and Source are in InDesign.
many of these services were new, arising from the beneficiaries.
coronavirus pandemic, the levels were above 2019
Source:
levels. Quality of care in 2021 was mixed relative
to 2019
• Notes
Clinic about
services. thisthere
In 2021, graph:
was a 3.4 million
Changes in our hospital quality indicators from 2019
increase in clinic services, or 0.9
• Data is in the datasheet. Make increase per in the datasheet.
updates to 2021 were mixed. FFS beneficiaries’ risk-adjusted
beneficiary (19.6 percent increase over 2020).
• I deleted the years from the x-axis and put in myhospital
Despite this large increase, the level remained
own. mortality rate increased slightly, while the
risk-adjusted hospital readmission rate improved.
• I had
below to manually
the 2019 level. draw tick marks and axis lines because they kept resetting when I changed any data.
Patient experience indicators declined.
• The data
Preliminary dashed line looked
for 2022 suggestok here,
that so I didn’t hand draw it.
outpatient
Quality of care in 2020 was difficult to assess due to
services
• I per FFS
can’t beneficiary
delete continued
the legend, so I’lltojust
increase
have to cropeffects
it out in InDesign.
of the coronavirus pandemic on beneficiaries
as hospital care continued to shift from inpatient
• Use direct selection tool to select items for modification. Otherwise
and providers. The if you use
results reflecthe black selection
temporary tool, they w
changes in
to outpatient settings and hospitals continued to
default when you change the data. the delivery of care and data limitations unique to the
acquire physician practices. The shift of services
public health emergency (PHE), so we did not use these
away • Useinpatient
from paragraph
carestyles (and object
is particularly styles) to format.
noteworthy
results to inform our conclusions about trends in the
because it coincided with a reduction in observation
quality of care provided to Medicare beneficiaries or
care. From 2019 to 2021, the number of outpatient
their relationship to Medicare payment adequacy.
observation visits per 1,000 Part B FFS beneficiaries
66 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
FIGURE
FIGURE
31-XX
–7 Title here....
From 2019 to 2021, FFS beneficiaries’ risk-adjusted all-condition mortality rate
during an inpatient stay or 30 days after discharge increased
12
Share of beneficiary deaths during an inpatient
10
8.8%
8 8.6% 8.6%
8.4%
8.1%
6
stay or 30 days after discharge (in percent)
4
Unadjusted mortality (data points not labeled)
2
Risk-adjusted mortality
0
2017 2018 2019 2020 2021
Note: FFS (fee-for-service). Data include hospitals paid under the inpatient prospective payment systems, critical access hospitals, and acute care
hospitals in Maryland and U.S. territories. The 2017–2020 risk-adjusted values differ slightly from what was presented in the March 2022 report
to the Congress because of a change in the baseline years and version of the 3M™ all-patient refined–diagnosis related group software for
calculating expected results. The 2019–2021 values are not connected because we cannot draw conclusions on the quality of care in 2020 due to
the effects of the coronavirus pandemic.
Source: MedPAC analysis of Medicare Provider Analysis and Review data.
Note: Note and Source are in InDesign.
Source:
This year we have updated the mortality and 2021 risk-adjusted mortality rate increased (that is,
Notes about this graph:
readmission risk-adjustment models to include the worsened) from 8.1 percent to 8.6 percent (Figure 3-7).
• Data
COVID-19 is in the
diagnosis, datasheet.
which Makeability
improves our updates
to in the From
datasheet.
2017 to 2019, the risk-adjusted mortality rate had
represent the acuity
• I deleted theand mix of
years patients
from and put in myimproved
receiving
the x-axis own. (that is, declined) by 0.7 percentage point.
hospital care in 2021. The 2021 patient experience Over the three-year period, unadjusted mortality rates
results•include
I had to manually
a full draw tick
year of survey marks
results andofaxis
instead a lines because
were they
relatively resetting
kept but
stable, when
expected I changed
mortality any data.
increased
partial•year.
The dashed line looked ok here, so I didn’t hand draw it.
because beneficiaries admitted to hospitals in recent
years tended to have more comorbidities and thus a
• I can’t delete the legend, so I’ll just have to crop it out in InDesign.
FFS beneficiaries’ risk-adjusted hospital mortality higher risk of mortality.
rate • Use direct selection tool to select items for modification. Otherwise if you use the black selection tool, they w
increased
default
In 2021, the when you
overall change
mortality rate the data.to rise
continued FFS beneficiaries’ risk-adjusted hospital
nationwide readmission rate improved
• Usedue to deaths from
paragraph COVID-19.
styles (and From
object 2019to
styles) to format.
2021, FFS beneficiaries’ unadjusted hospital mortality Many factors related to the coronavirus pandemic
rate (death during a hospital stay or 30 days after affected hospitalization rates, including both greater
discharge) increased from 8.4 percent to 11.5 percent demand for beds for patients diagnosed with COVID-19
(data points are not labeled). During that time, the and lower demand for beds because some patients
Report to the Congress: Medicare Payment Policy | March 2023 67
FIGURE
FIGURE
3– 8 Title here....
From 2019 to 2021, FFS Medicare beneficiaries’ all-condition
1-XX
risk-adjusted hospital readmission rate improved
18
Share of beneficiaries readmitted for an inpatient
15.9%
16
15.8% 15.7%
14 15.1% 14.9%
12
10
stay within 30 days after an initial stay (in percent)
8
6
4
Unadjusted unplanned admissions (data points not labeled)
2 Risk-adjusted unplanned admissions
0
2017 2018 2019 2020 2021
Note: FFS (fee-for-service). Data include FFS Medicare beneficiaries ages 65 or older and hospitals paid under the inpatient prospective payment
systems, critical access hospitals, and acute care hospitals in Maryland and U.S. territories. The 2017–2020 unadjusted values differ slightly
from values presented in the March 2022 report to the Congress because of minor changes to the measure calculation. The 2017–2020 risk-
adjusted values differ because of a change in the baseline years and version of the 3M™ all-patient refined–diagnosis related group software for
calculating expected results. The 2019–2021 values are not connected because we cannot draw conclusions on the quality of care in 2020 due to
the effects of the coronavirus pandemic.
Source: MedPAC analysis of Medicare Provider Analysis and Review data.
Note: Note and Source are in InDesign.
Source:
avoidedNotes about
hospitals due othis
fears graph:
of infection. Between Hospitals collect Hospital Consumer Assessment of
2019 and 2021, the unadjusted rate of readmissions (FFS
• Data is in the datasheet. Make updates in the datasheet. Healthcare Providers and Systems® (H–CAHPS®)
Medicare beneficiaries over age 65 readmitted within surveys from a sample of admitted patients, which CMS
• I deleted the years from the x-axis
30 days after discharge) increased by 0.1 percentage and put in my own.
uses to calculate results for 10 measures of patient
13
point • I had
(data not to manually
labeled) draw
(Figure 3-8).tick
Themarks
rate ofandrisk-axis lines because they
experience keptinresetting
included hospitals’when
overallI ratings.
changed any data.
adjusted readmissions decreased (that is, improved) The H–CAHPS measures key components of quality
• The dashed line looked ok here, so I didn’t hand draw it.
by 0.8 percentage point to 14.9 percent because by assessing whether something that should happen
• I can’t
beneficiaries delete to
admitted thehospitals
legend,inso I’ll just
recent have to cropduring
years it outainhospital
InDesign.
stay (such as clear communication)
tended•toUsehave moreselection
direct comorbidities
tool toand thus aitems
select higher actually happened
for modification. Otherwiseor if
how often
you useitthe
happened.
black In 2021, 72
selection tool, they w
expected rate of
default readmission.
when you change the data. percent of surveyed patients rated their overall hospital
experience a 9 or 10 on a 10-point scale, which is a 1
Patient• Use paragraph
experience styles
results (and object styles) to format.
declined percentage point decrease from 2019. Communication
Between 2019 and 2021, hospital patient experience with nurses, communication with doctors, and receipt
measures remained high but most declined by 1 of discharge information had the highest scores, with
percentage point to 4 percentage points (Table 3-2). at least 80 percent of surveyed patients answering
68 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
TABLE
Hospital patient experience measures declined from 2019‒2021
3–2
Percentage
point change,
®
H‒CAHPS measure 2017 2018 2019 2020 2021 2019–2021
Share of patients rating the hospital a 73% 73% 73% 72% 72% –1
9 or 10 out of 10
Share of patients who would definitely 72 72 72 71 70 –2
recommend the hospital
Share of patients giving top ratings for:
Communication with nurses 80 81 81 80 80 –1
Communication with doctors 82 81 82 81 80 –2
Responsiveness of hospital staff 70 70 70 67 66 –4
Communication about medicines 66 66 66 63 62 –4
Cleanliness of hospital environment 75 75 76 73 73 –3
Quietness of hospital environment 62 62 62 63 62 0
Understanding their care when they 53 53 54 52 52 –2
left the hospital (care transitions)
Share of patients who received 87 87 87 86 86 –1
discharge information
Note: H‒CAHPS® (Hospital Consumer Assessment of Healthcare Providers and Systems®). H‒CAHPS is a standardized 32-item survey of patients’
evaluations of hospital care. The survey items are combined to calculate measures of patient experience for each hospital. The H‒CAHPS measures
included in the table are “top-box,” or the most positive, response to H‒CAHPS survey items. Each year’s results are based on a sample of surveys of
hospitals’ patients from January to December. Results in 2020 include only surveys from patients discharged July to December 2020 rather than
the customary full year.
Source: CMS summary of H‒CAHPS public report of survey results tables.
with the most positive response. From 2019 to 2021, program (HVIP)—that balances rewards and penalties
responsiveness of hospital staff and communication and has the potential to drive further improvement
about medicines dropped by 4 percentage points, in hospital quality (Medicare Payment Advisory
and cleanliness of hospital environment dropped by 3 Commission 2019). Initially, the HVIP could incorporate
percentage points. In 2021, the care-transition measure existing quality measure domains such as readmissions,
continued to get the lowest score, with only 52 percent mortality, spending, patient experience, and hospital-
of surveyed patients responding with “Strongly Agree” acquired conditions (or infection rates). A key feature
that they understood their care plan when they left the of the Commission’s HVIP design is that it accounts
hospital. for differences in providers’ patient populations by
incorporating a peer-grouping methodology. Quality-
Need for a redesign of hospital quality based payments would be distributed to hospitals
payment programs separated into peer groups, defined by their share of
In March 2019, the Commission recommended that the beneficiaries who have full dual eligibility for Medicare
Congress replace Medicare’s current hospital quality and Medicaid (as a proxy for income). The grouping
programs (including the penalty-only programs) with of hospitals into peer groups that serve similar
a single, outcome-focused quality-based payment populations would make payment adjustments more
program for hospitals—a hospital value incentive equitable than existing quality payment programs.
Report to the Congress: Medicare Payment Policy | March 2023 69
FIGURE
FIG RE Title here....
1-XX
3–9 IPPS hospitals’ all-payer operating margin reached a
record high in 2021, despite declines in federal relief funds
20
15
All-payer operating margin 10 8.7%
6.4%
5.9% 5.9%
(in percent)
5.3%
5 7.2%
1.9%
0
–5
2017 2018 2019 2020 2021
Aggregate, including relief funds Excluding relief funds Middle half of hospitals
Note: IPPS (inpatient prospective payment systems). Hospitals’ margins are calculated as aggregate payments minus aggregate costs, divided
by aggregate payments. “All-payer” margin includes payments from all payers. The “operating” margin is limited to patient care and other
operating revenue, and in 2020 and 2021 these margins are reported with and without federal relief funds (Provider Relief Fund payments and
forgiven loans from the Paycheck Protection Program). Data are for IPPS hospitals that had a cost report with a midpoint in the fiscal year and
that was complete as of our analysis.
Source: MedPAC analysis of hospital cost reports.
Note: Note and Source are in InDesign.
Source:
Hospitals’
Notes access
about to capital
this strengthened in
graph: and to 7.2 percent without federal relief funds, both
2021 but was mixed in 2022 of which were higher than the prior all-time high in
• Data is in the datasheet. Make updates in the datasheet. 2019 (Figure 3-9).14 The increase in IPPS hospitals’ all-
Hospitals’ access to capital strengthened in 2021, with
IPPS • I deleted the
hospitals’ years
all-payer from the
operating x-axis and
margin reachingput in my payer
own.operating margin occurred despite a decrease
a record high despite declining in federal relief funds: In their fiscal year 2021 cost-
• I had to manually drawfederal relief funds.
tick marks and axis lines because they kept resetting when I changed any data.
However, margins continued to vary substantially reporting period, hospitals reported receiving about
• The dashed
across hospitals.
line looked ok here, so I didn’t hand draw
$18 it. in these funds, down from $35 billion
billion
15
• I can’t delete the legend, so I’ll just have to cropinit out 2020.in In other words, the federal relief funds
InDesign.
Preliminary 2022 all-payer operating margin data were that hospitals received in 2021 more than offset the
mixed•relative
Use direct selection tool
to prepandemic to select
levels, items for modification. Otherwise if you use the black selection tool, they w
but hospitals additional coronavirus pandemic–related expenses
default
continued to when you change
have strong access tothe data.
bond markets. that were not covered by the higher patient revenues
• Use paragraph styles (and object styles) to format. associated with COVID-19. Rather, the increase in
Hospitals’ all-payer operating margin reached a the operating margin of over 3 percentage points
record high in 2021
resulted from hospitals’ operating revenues growing
In 2021, IPPS hospitals’ all-payer operating margin more than their costs: Operating revenue increased
increased to 8.7 percent with federal relief funds
70 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
TABLE
In 2021, IPPS hospitals’ all-payer operating margins continued to vary across
3–3 hospital groups, including an all-time high among for-profit hospitals
2020 2021
With Without With Without
Hospital group 2017 2018 2019 relief funds relief funds relief funds relief funds
All IPPS 5.9% 5.9% 6.4% 5.3% 1.9% 8.7% 7.2%
Ownership
For profit 10.5 11.4 12.2 12.6 10.4 15.1 13.9
Nonprofit 5.9 5.5 6.1 4.7 1.2 8.2 6.8
Location
Metropolitan (urban) 6.0 6.1 6.6 5.3 2.0 8.6 7.3
Rural micropolitan 4.9 3.9 5.2 6.2 1.9 9.2 6.8
Other rural 2.1 0.2 0.7 3.4 –1.5 7.6 3.0
Teaching and DSH
Both 5.7 5.8 6.2 4.8 1.4 8.4 6.9
DSH only 5.5 5.6 6.3 6.2 2.8 8.9 7.3
Teaching only 8.8 8.7 7.7 6.0 4.1 7.7 6.7
Neither 9.0 9.1 10.1 8.4 6.0 13.5 11.8
Note: IPPS (inpatient prospective payment systems), DSH (disproportionate share hospital). Hospitals’ margin is calculated as aggregate payments minus
aggregate costs, divided by aggregate payments. “All-payer operating margin” includes patient care and other operating revenue from all payers,
and, for 2020 and 2021, is reported with and without reported federal relief funds (Provider Relief Fund payments and Paycheck Protection Program
forgiven loans). Metropolitan (urban) counties contain an urban cluster of 50,000 or more people; rural micropolitan counties contain a cluster of
10,000 to 50,000 people; all other counties are classified as “other rural.” Data are for IPPS hospitals that had a cost report with a midpoint in the
specified fiscal year and that were complete as of our analysis.
Source: MedPAC analysis of hospital cost reports and census geographic files.
over 11 percent, while costs increased by only about 7 latter known as disproportionate share hospitals
percent.16 Several large hospital systems highlighted (DSHs)) (Table 3-3). In contrast, the operating margin
the growth in inpatient acuity as contributing to their continued to be lower among hospitals in rural
improved operating margin. nonmicropolitan areas. However, rural hospitals
received targeted federal relief funds, so the difference
Within hospitals’ aggregate all-payer operating in the all-payer operating margin between rural and
margin, there continued to be significant variation: urban hospitals was smaller than prepandemic levels.
The 2021 operating margin ranged from 0.8 percent to
14.9 percent among the middle half of IPPS hospitals Preliminary 2022 all-payer operating margin data
(Figure 3-9, data not labeled). While there was variation are mixed relative to prepandemic levels
within each group of hospitals, in aggregate, the
Preliminary data from several large hospital systems
operating margin continued to be higher among for-
suggest hospitals’ all-payer operating margin declined
profit hospitals and those that were neither teaching
during the first half of 2022 relative to the record high
nor receiving disproportionate share payments (the
Report to the Congress: Medicare Payment Policy | March 2023 71
margins in 2021. Because 2021 was atypical, we also access the bond market. During the start of the
compare these large systems’ 2022 all-payer margins coronavirus pandemic in spring 2020, investors
with their prepandemic (2019) operating margins. demanded a larger premium to hold hospital bonds,
Our sample of partial year data for 2022 is limited to reaching a peak of 3.5 percentage points above the
five large systems that represent about 17 percent yield on treasury bonds. This peak was well above the
of all IPPS hospitals. From the fiscal year ending in premium over the past 10 years, which generally ranged
June 2019 to the fiscal year ending in June 2022, two from 1 percentage point to 2.5 percentage points above
of the largest nonprofit systems reported a decline the yield on treasury bonds. By the start of fiscal year
in operating margins of 2 percentage points to 4 2021, hospitals’ risk premium to access bonds had
percentage points (Ascension 2022a, Ascension 2020, declined to 2.5 percentage points above the yield on
Trinity Health 2022a, Trinity Health 2020). In contrast, treasury bonds. Throughout most of 2021 and 2022,
during the first nine months of 2022 compared with hospitals’ risk premium to access bonds continued to
the first nine months of 2019, two of the three largest decline, falling to 1 percentage point above the yield
for-profit systems reported an increase in operating on treasury bonds by the end of fiscal year 2022 (S&P
margins of 2 percentage points to 4 percentage points Global 2022).
(HCA Healthcare 2022, HCA Healthcare 2020, Tenet
Health 2022, Tenet Health 2020); the third reported While investor interest in bonds remained strong, by
a 2022 operating margin similar to its 2019 margin the end of 2022 S&P Global Ratings downgraded about
(Community Health Systems 2022, Community Health 10 percent of nonprofit hospital bonds (S&P Global
Systems 2020). Aggregating the data from these five Ratings 2022). At the start of fiscal year 2023, the
systems, all-payer operating margins remained positive ratings agencies reported a stable outlook for about 80
and about equal to 2019 levels.17 There is still a material percent of nonprofit hospitals, a negative outlook for
level of uncertainty regarding labor costs and overall about 15 percent of nonprofit hospitals, and a positive
profitability in the fourth quarter of 2022. outlook for about 5 percent of nonprofit hospitals
(Moody’s Investors Service 2022, S&P Global Ratings
A few factors influenced these large hospital systems’ 2022).
lower operating margins in 2022 relative to those in
2021. Inflation was higher than expected and led to IPPS hospitals’ Medicare margin improved
higher operating expenses. Hospital systems also in 2021 and was near break-even for
cited a combination of ongoing workforce shortages, relatively efficient hospitals
high labor costs, a reduction of provider relief funds, From 2020 to 2021, IPPS hospitals’ Medicare margin
and declining patient acuity as some reasons for the increased, with the median Medicare margin among
downward fiscal pressure on margins. The decline relatively efficient hospitals becoming positive when
in patient acuity in 2022 was attributed to declining including Medicare’s share of federal relief funds and
COVID-19 volume, reversing a prior trend of higher increasing from negative to break-even when excluding
acuity due to high COVID-19 volume.18 However, these funds.
several hospital systems reported improvements
in financial performance for the second and third IPPS hospitals’ Medicare margin remained
quarters of 2022. Those improvements were negative in 2021 but increased above
attributed to near-term favorable trends in patient prepandemic levels
volume, fewer COVID-19 disruptions, and a decline In 2021, IPPS hospitals’ aggregate Medicare margin
in the use of contract labor (Ascension 2022b, across hospital service lines remained negative but
CommonSpirit 2022, Community Health Systems increased above prepandemic levels, even before
2022, Trinity Health 2022b). including any federal relief funds (Figure 3-10).19
Specifically, IPPS hospitals’ Medicare margin increased
Hospitals continued to have strong access to to –6.2 percent in 2021 when including Medicare’s
bond markets share of federal relief funds—a recent high.20 Exclusive
In both fiscal years 2021 and 2022, hospitals have of these funds, the Medicare margin was –8.2 percent,
continued to pay a relatively low risk premium to which was still higher than prepandemic levels. The
72 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
FIGURE
FIG RE Title here....
1-XX
3–10 IPPS hospitals’ Medicare margin remained negative in 2021
but increased to above prepandemic levels
10
5
Aggregate Medicare margin
0
–6.2%
–5
–8.2%
–8.5%
–9.9% –9.3%
(in percent)
–10 –8.2%
–12.3%
–15
–20
–25
2017 2018 2019 2020 2021
Aggregate, including relief funds Excluding relief funds Middle half of hospitals
Note: IPPS (inpatient prospective payment systems). “Relief funds” refers to Provider Relief Fund payments and forgiven loans from the Paycheck
Protection Program, as recorded on hospitals’ cost reports; Medicare’s share of these funds was calculated using fee-for-service Medicare’s share
of 2019 all-payer operating revenue. Hospitals’ “Medicare margin” is calculated as aggregate Medicare payments minus aggregate allowable
Medicare costs, divided by aggregate Medicare payments. Payments and costs include multiple hospital service lines (including inpatient,
outpatient, swing bed, skilled nursing, rehabilitation, psychiatric, and home health services) as well as direct graduate medical education and
uncompensated care payments. Data are for IPPS hospitals that had a cost report with a midpoint in the fiscal year and that was complete as of
our analysis.
Source: MedPAC analysis of hospital cost reports.
Note: Note and Source are in InDesign.
Source:
Notes about this graph:
• Data is in the datasheet. Make updates in the datasheet.
increase in IPPS hospitals’ Medicare margin from 2020 percent among the middle half of IPPS hospitals
• Iof 2
to 2021 deleted the years
percentage pointsfrom the 4x-axis
(about and put in my own.
percentage (Figure 3-10, data not labeled). While there was
points
• Ibefore including
had to manuallyrelief
drawfunds)
tickresulted
marks and from axis linesvariation
becausewithin each group
they kept of when
resetting hospitals,
I in aggregate,
changed any data.
hospitals’ Medicare revenues growing while their costs the Medicare margin continued to be higher—and
• The dashed line looked ok here, so I didn’t handpositive—at
held relatively steady. In other words, in aggregate,
draw it. for-profit hospitals and hospitals in small
• I can’tpayment
the Medicare delete the legend,
increases so I’llthe
during justPHE
have
moreto croprural
it outcommunities
in InDesign.(Table 3-4, p. 74). In contrast,
than •
offset
Usehospitals’ additional
direct selection pandemic-related
tool to select items for costs the Medicare
modification. marginifcontinued
Otherwise to be
you use the lowerselection
black among tool, they w
and increased the share of hospitals that had a positive hospitals that were not disproportionate share
default when you change the data.
Medicare margin. (For a description of the Medicare hospitals.
• Use
payment paragraph
increases, styles
see the text(and
box in object
our Marchstyles)2022
to format.
In 2021, aggregate IPPS payments increased while
report to the Congress, p. 89.)
hospitals’ aggregate inpatient costs decreased slightly
Within hospitals’ aggregate Medicare margin, there In 2021, aggregate IPPS payments to hospitals for FFS
continued to be significant variation: The 2021 Medicare beneficiaries’ inpatient stays increased 3.4
Medicare margin ranged from –15.2 percent to +8.1 percent to $107.9 billion (Figure 3-11, p. 75, left panel).
Report to the Congress: Medicare Payment Policy | March 2023 73
TABLE
In 2021, IPPS hospitals’ Medicare margin continued to vary across
3–4 hospital groups, including higher margins at for-profit and rural hospitals
2020 2021
With Without With Without
Hospital group 2017 2018 2019 relief funds relief funds relief funds relief funds
All IPPS –9.9% –9.3% –8.5% –8.2% –12.3% –6.2% –8.2%
Ownership
For profit –2.2 –0.3 1.3 4.3 1.6 5.3 3.7
Nonprofit –11.1 –10.6 –10.0 –10.3 –14.8 –8.2 –10.2
Location
Metropolitan (urban) –10.1 –9.5 –8.8 –8.7 –12.8 –6.6 –8.5
Rural micropolitan –8.3 –7.1 –6.1 –3.7 –8.5 –2.6 –5.8
Other rural –5.6 –5.2 –2.5 1.6 –4.0 4.9 –0.8
Teaching and DSH
Both –8.7 –8.4 –7.8 –7.7 –11.8 –5.8 –7.8
DSH only –11.2 –10.3 –9.1 –7.9 –12.2 –5.7 –8.0
Teaching only –14.3 –12.0 –11.7 –14.4 –16.9 –11.0 –12.5
Neither –17.2 –15.3 –14.3 –13.9 –17.0 –10.8 –13.3
Note: IPPS (inpatient prospective payment systems). Hospitals’ margin is calculated as aggregate payments minus aggregate allowable costs, divided
by aggregate payments. Hospitals’ “Medicare margin” is calculated as aggregate Medicare payments minus aggregate allowable Medicare costs,
divided by aggregate Medicare payments. Payments and costs include multiple hospital service lines (including inpatient, outpatient, swing
bed, skilled nursing, rehabilitation, psychiatric, and home health services) as well as direct graduate medical education and uncompensated care
payments. For 2020 and 2021, the margin is reported with and without reported federal relief funds (Provider Relief Fund payments and Paycheck
Protection Program forgiven loans). Metropolitan (urban) counties contain an urban cluster of 50,000 or more people; rural micropolitan counties
contain a cluster of 10,000 to 50,000 people; all other counties are classified as “other rural.” Data are for IPPS hospitals that had a cost report with a
midpoint in the specified fiscal year and that were complete as of our analysis.
Source: MedPAC analysis of hospital cost reports and census geographic files.
This increase in payments occurred despite a decrease resulted from: (1) a 2.4 percent annual update to
in FFS beneficiaries because there was a larger growth the inpatient operating base payment rate; (2) a
in Medicare payments per stay, which rose 10.3 percent 0.5 percent statutory increase to the inpatient
to about $15,600 (Figure 3-11, right panel). operating base rate (from phasing out adjustments
that were put in place in 2018 to recoup prior
The 10.3 percent growth in IPPS payments per stay in coding-induced overpayments); and (3) a 1.1 percent
2021 resulted primarily from: update to the inpatient capital base rate. Because
• Annual update to the IPPS base payment rates the inpatient operating rate is about 97 percent
and statutory increase. In 2021, the net annual of total IPPS base rates, the net update was 2.8
update to IPPS base payment rates—including a percent ((93 percent × 2.9 percent) + (7 percent × 1.1
statutory increase—was 2.8 percent. This increase percent)).
74 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
FIGU R
FIGURE
Title here....
3 –11
1-XX In 2021, aggregate IPPS payments for FFS beneficiaries’ inpatient stays remained
below the prepandemic level while payments per stay continued to rise rapidly
150 20
Aggregate IPPS payments (in billions)
IPPS payments per stay (in thousands)
$111.4 $111.4 $15.6
$110.5
$104.3 $107.9 $14.2
15
$13.0
$12.6
100 $12.2
10.3%
3.4% 10
50
5
0 0
2017 2018 2019 2020 2021 2017 2018 2019 2020 2021
Note: IPPS (inpatient prospective payment systems), FFS (fee-for-service). IPPS payments exclude payments for uncompensated care because these
are not payments for FFS Medicare beneficiaries’ inpatient stays. Payments per stay are per transfer-adjusted stay.
Source: MedPAC analysis of Medicare Provider Analysis and Review data for fiscal years 2017 to 2021.
• Growth in case mix. In 2021, there was a 3.4 percent Meanwhile, between 2020 and 2021, IPPS hospitals’
Note: Note and Source are in InDesign.
increase of reported inpatient case mix, net of aggregate costs for inpatient services fell slightly. This
changes from annual updates to relative weights.
Source: decrease was the combination of two factors. First,
These weights do not consider patients’ COVID-19 the number of FFS Medicare inpatient stays declined.
status. Second, IPPS hospitals were able to constrain the
Notes about this graph:
growth in costs per inpatient stay to slightly below
• Increases
• Data in
is Medicare paymentsMake
in the datasheet. duringupdates
the PHE.in the the
datasheet.
increase in input prices and average case mix.
We estimate that the suspension of the 2 percent
• I deleted the years from the x-axis and put in myThis own. constraint in cost growth is similar to prior years,
sequestration in the Medicare program’s share of except for 2020, and indicates that hospitals coded
• I had
FFS to which
payments, manuallybegan
drawontick marks
May 1, 2020,and
andaxis linespatients
because they kept resetting when I changed any data.
more extensively, improved productivity, or
extended
• throughline
The dashed April 2022, ok
looked raised IPPSso I
here, payments
didn’t handboth.
drawAs it.
the increase in costs per inpatient stay was
per stay by 1.1 percent in 2021.21 In addition, we slightly lower than the decline in the number of FFS
• I can’t delete the legend, so I’ll just have to crop it out in InDesign.
estimate that the mandated 20 percent increase Medicare inpatient stays, IPPS hospitals’ aggregate
in•the
Use direct weight
resource selection
fortool to select
inpatient staysitems
whenfor modification.
inpatient Otherwise if you
costs declined use the black selection tool, they w
slightly.
defaulthave
patients when a you change
COVID-19 the data.
diagnosis increased
2021 IPPS payments per stay by an additional 1.2 In 2021, uncompensated care payments held steady In
• Use paragraph styles (and object styles) to format.
percent. We also estimate that add-on payments 2021, uncompensated care payments—payments that
for new COVID-19 technologies increased 2021 the Medicare program makes to help cover hospitals’
IPPS payments per stay by an additional 1 percent. costs of bad debt and charity care—held steady at near
Report to the Congress: Medicare Payment Policy | March 2023 75
FIGU R
FIGURE
Title here....
31-XX
–12 In 2021, OPPS aggregate payments for outpatient services and
payments per FFS beneficiary rose above prepandemic levels
OPPS payments per beneficiary (in thousands)
100 3
Aggregate OPPS payments (in billions)
80 $2.2
$66.9
$63.3 $66.2 $60.0 $2.0
$1.9 $1.9
$59.1 2 $1.8
60
16.5%
11.4%
40
1
20
0 0
2017 2018 2019 2020 2021 2017 2018 2019 2020 2021
Note: OPPS (outpatient prospective payment system). The data include all OPPS payments (including, but not limited to, general acute care
hospitals). Data reported by calendar year.
Source: MedPAC analysis of Medicare outpatient claims data and the Medicare Trustees report.
$8.3 billion
Note: Note andand therefore
Source are in did not materially contribute
InDesign. not affect the payment rates of separately payable
to the
Source:
increase in hospitals’ Medicare margin in 2021.22 drugs and devices, which are based on average
acquisition costs and represent 26 percent of OPPS
In 2021, OPPS aggregate payments and payments per
payments.23 Therefore, the average effect of the
beneficiary increased rapidly, outpacing growth in
Notes about this graph: annual update on spending across OPPS services
hospitals’ outpatient costs In 2021, OPPS payments
• Medicare
Data is inbeneficiaries’
the datasheet. Make services was 1.8 percent ((74 percent × 2.5 percent) + (26
updatesatin the datasheet.
for FFS outpatient percent × 0 percent)).
general• IACHs increased
deleted to $66.9
the years billion,
from the which
x-axis wasput in my own.
and
slightly above prepandemic levels despite a decrease in Growth in service volume. In 2021, the volume
• I had to manually draw tick marks and axis lines• because they kept resetting when I changed any data.
the number of FFS Medicare beneficiaries (Figure 3-12, of OPPS services per beneficiary raised OPPS
• TheMeanwhile,
left panel). dashed line OPPSlooked ok here,
payments so I Bdidn’t
per Part FFS hand draw it. per FFS beneficiary by 13.5 percent.
payments
beneficiary increased
• I can’t delete to
theabout $2,200,
legend, so aI’llsharp increase
just have Thisinincrease
to crop it out was driven by a general increase
InDesign.
from the 2020 level and above the prepandemic 2019 in all types of hospital outpatient department
• Use direct selection tool to select items for modification. Otherwise if you use the black selection tool, they w
level (Figure 3-12, right panel). (HOPD) services and by the provision of 7.7 million
default when you change the data. COVID-19 vaccine administrations and testing for
The 16.5 percent
• Use growth in
paragraph OPPS
styles payments
(and objectper FFS to format. COVID-19.
styles)
Medicare beneficiary in 2021 resulted primarily from:
• Decline in complexity. In 2021, OPPS payments per
• Annual update to OPPS conversion factor. In 2021, service fell 1.8 percent due to the mix of outpatient
the annual update to the OPPS conversion factor services, measured by the OPPS relative weights
was 2.5 percent. However, the OPPS update does
76 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
of the services. This measure decreased because funds (Table 3-5, p. 78).25 These findings are consistent
of a sharp increase in relatively low-complexity with data over the last several years showing relatively
services, especially administration of the COVID-19 efficient hospitals approximately breaking even on
vaccines and testing for COVID-19. Medicare.26 (As in prior years, we identified relatively
efficient hospitals as those that were never in the worst
• Continued growth in spending on separately third on any quality or cost metrics during the prior
payable drugs. Payments for separately payable three years (we used 2017, 2018, and 2019 to limit the
drugs grew 9.8 percent per beneficiary. Separately effect of the start of the pandemic) and consistently
paid drugs are about 26 percent of total OPPS performed in the top third of either costs or mortality
spending, so this increase in drug spending boosted (see text box, p. 79); however, to limit the effect of the
OPPS spending per beneficiary by 2.6 percent. start of the pandemic on these measures and hospitals’
• Increases in Medicare payments during the PHE. different cost-reporting periods, we used 2017, 2018,
We estimate that the suspension of the 2 percent and 2019 to identify relatively efficient hospitals and
sequestration in Medicare’s share of FFS payments, then looked at their performance in 2021.)
which began on May 1, 2020, and extended through In 2021, the relatively efficient hospitals’ lower costs
April 2022, raised OPPS payments per beneficiary per inpatient stay (91 percent of the national median)
by 0.6 percent in 2021.24 allowed them to generate better Medicare margins
Meanwhile, hospitals’ outpatient aggregate costs and than the comparison group. The relatively efficient
cost per beneficiary increased but at a slower rate. group also had better patient satisfaction, with 71
The increase in costs reflects the large increase in percent of H–CAHPS respondents rating the hospital
outpatient services per beneficiary, a small increase in a 9 or 10 in 2020, compared with 68 percent for other
input prices, a small increase in the cost of separately hospitals. The relatively efficient hospitals (those that
payable drugs and devices, and a decrease in the had relatively good prepandemic quality metrics)
resource requirements per OPPS-covered service. One continued to have lower risk-adjusted median mortality
driver of the decreased resource requirements was the and readmission rates than other hospitals during
large volume of COVID-19 vaccine administrations and the pandemic. Among our sample of 284 relatively
COVID-19 sample collections (7.7 million), which are efficient hospitals in 2021, mortality was 7 percentage
low-complexity services. One reason why hospitals’ points below the national median and readmission
Medicare outpatient costs grew more slowly than rates were 4 percentage points below the national
Medicare payments in 2021 is that the suspension of median—consistent with comparisons in 2017 to 2019.
the 2 percent sequestration on Medicare program These results suggest that relatively efficient and
payments was in effect for all of calendar year 2021 other hospitals’ mortality and readmission metrics, on
compared with only a portion of 2020, which increased average, were equally affected by the pandemic.
payments without affecting costs. A second possible As in past years, relatively efficient hospitals
explanation for why hospitals’ outpatient costs grew were spread across the country and represented
more slowly than Medicare payments is that the costs diverse categories of hospitals, including teaching,
incurred when providing COVID-19 vaccines and nonteaching, rural, urban, for-profit, and nonprofit
taking sample collections were smaller than the OPPS hospitals, as well as hospitals serving large shares
payments for those services. of low-income patients. On average, the shares of
Medicare and Medicaid patients are similar in both
In 2021, relatively efficient hospitals’ median
groups. While most types of hospitals were represented
Medicare margin was positive after including
Medicare’s share of federal relief funds in the efficient group, a disproportionate share of
relatively efficient hospitals had relatively high volumes
In 2021, the median Medicare margin among the 15
of admissions. Volume primarily affects our efficiency
percent of IPPS hospitals we identified as relatively
measures in two ways. First, higher-volume hospitals
efficient remained at 1 percent when including
tended to have lower risk-adjusted mortality. Second,
Medicare’s share of federal relief funds and increased
we require some consistency of results over three years
from –3 percent in 2020 to break-even excluding these
and remove from the efficient group any hospital that
Report to the Congress: Medicare Payment Policy | March 2023 77
TABLE
3–5 Performance of relatively efficient hospitals in 2021
Type of hospital
Relatively
Relative performance measure efficient Other
Number of hospitals 284 1,672
Share of hospitals in our study sample 15% 85%
Historical performance, 2017–2019 (percent of national median)
Mortality rate 89% 101%
Readmission rate 93 102
Standardized Medicare costs per stay 90 103
Performance metrics, 2021 (percent of national median)
Mortality rate 93% 101%
Readmission rate 96 101
Standardized Medicare costs per stay 91 102
Share of patients rating the hospital a 9 or 10 (out of 10), 2021 71 68
Median Medicare margin, 2021
Medicare margin excluding relief funds 0% −7%
Medicare margin with relief funds 1 −4
Median all-payer total margin 11 9
Note: “Relatively efficient hospitals” and “other hospitals” were identified based on their mean performance during 2017–2019 relative to the median
hospital’s performance during those years. We removed hospitals with a low share of Medicaid patient days reported on cost reports (the bottom
10 percent of hospitals) and hospitals in markets with high service use (top 10 percent of hospitals) due to concerns that socioeconomic conditions
and aggressive treatment patterns can influence unit costs and risk-adjusted quality metrics. Data differ slightly from the data presented in our
March 2022 report because we limit this set of data to providers that had 2021 cost report data. “Mortality rate” is the risk-adjusted rate of mortality
within an inpatient stay through 30 days after the stay. “Readmission rate” is the risk-adjusted rate of readmission within 30 days of an inpatient
stay. “Standardized Medicare costs per stay” is standardized for area wage rates, case-mix severity, prevalence of outlier and transfer cases, interest
expense, low-income shares, and teaching intensity. “Share of patients rating the hospital a 9 or 10 (out of 10)” is based on Hospital Consumer
Assessment of Healthcare Providers and Systems survey data collected from patients discharged July to December of 2021.
Source: MedPAC analysis of cost report and claims-based quality data from CMS.
performed in the bottom third on any metric in a single hospitals tend to have lower costs, one might expect
year.27 Thus, random variation in smaller hospitals them to be in the efficient group. However, between 14
may make them more likely to be excluded from our percent and 15 percent of both for-profit and nonprofit
efficient group. The efficient group also tends to have hospitals were deemed relatively efficient. The
lower shares of low-income patients.28 factor that separates the relatively efficient hospitals
from other low-cost hospitals is that they perform
This year, as in past years, we have found that for-profit relatively well on both quality and costs. While for-
hospitals have been able to break even or generate profit hospitals tended to have lower costs, nonprofit
small profits on Medicare patients (Medicare Payment hospitals tended to perform slightly better on our
Advisory Commission 2022a). Given that for-profit quality metrics.
78 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
Identifying relatively efficient hospitals
T
he Commission follows two principles when • Risk-adjusted mortality rates were not among the
identifying a set of efficient providers. First, worst third in any year.
the providers must do relatively well on cost
and quality metrics. Second, the performance has • Risk-adjusted readmission rates were not among
to be consistent, meaning that the provider cannot the worst third in any year.
have poor performance on any metric over the past
three years. In the hospital sector, the variables • Standardized costs per inpatient stay were not
we use to identify relatively efficient hospitals are among the worst third in any year.
hospital-level mortality rates (risk-adjusted, all-
condition mortality during an inpatient stay through • Risk-adjusted mortality or standardized costs per
30 days after discharge), readmission rates (risk- stay were among the best one-third of all hospitals
adjusted, all-condition readmission rates within in all years.
30 days after an initial stay), and standardized
The objective was to identify a sample of hospitals
inpatient Medicare costs per case. Our assessment
that consistently performed at an above-average
of efficiency is not in absolute terms but, rather,
level on at least one measure (cost or mortality)
relative to a comparison group of other hospitals
and that always performed reasonably well on all
paid under Medicare’s inpatient prospective
measures. Because we screen out hospitals that have
payment systems (IPPS).29
few Medicaid patients or have poor performance
in a single year, our methodology does not seek
Categorizing hospitals as relatively efficient
to identify all efficient hospitals, only a subsample
We assigned IPPS hospitals to the relatively efficient of relatively efficient hospitals. The rationale for
group or the control group according to each this methodology and the details of computing the
hospital’s performance relative to the national various measures are discussed in our March 2011
median on a set of risk-adjusted cost and quality report (Medicare Payment Advisory Commission
metrics for the three years prior to the most 2011). As a secondary check on hospital quality,
recent cost report year. We then examined the we use the Hospital Consumer Assessment of
performance of the two hospital groups in the most Healthcare Providers and Systems survey to require
recent cost report year. that at least 60 percent of the hospital’s patients
rated it a 9 or 10 on a 10-point scale (in the year
Hospitals were identified as relatively efficient if prior to the performance period).30 ■
they met four criteria in each of the three prior
years:
Hospitals’ Medicare margin for 2023 is projected and costs from the most recent year of complete data
to decline relative to 2021 (2021) and policy, inflation, and coronavirus pandemic–
We project that IPPS hospitals’ aggregate Medicare related changes that took place in 2022 and are
margin for 2023 will decline relative to 2021. anticipated in 2023.
Specifically, we project that their Medicare margin in
The following are key drivers of our projected lower
2023 will be approximately –10 percent, similar to the
Medicare margin in 2023 relative to 2021:
level in 2017. Among relatively efficient IPPS hospitals,
we project that the median Medicare margin in 2023 • Hospitals’ input prices growing faster than CMS’s
will be modestly below break-even, near prepandemic forecast. In 2022, CMS underestimated the growth
levels. These projections are based on actual payments in hospitals’ input prices when it set the annual
Report to the Congress: Medicare Payment Policy | March 2023 79
update for IPPS and OPPS payment rates (Table have any corresponding Medicare costs, all declines
3-6). Using data available as of the time of the 2022 in uncompensated care payments decrease
final rule (published in 2021), CMS forecast that hospitals’ Medicare margins. (On the other hand, a
general ACHs’ input prices for a market basket of decrease in the uninsured rate generally increases
operating inputs would increase by 2.7 percent hospitals’ all-payer margin.)
from 2021 to 2022. However, CMS’s latest forecast
(with historical data through the second quarter of We anticipate that these factors which reduce net
calendar year 2022) suggests that input prices in revenue growth will be partially offset by other factors
fiscal year 2022 grew 5.7 percent (3.0 percentage that reduce cost growth, including reductions in
points higher than initially forecast). There is hospitals’ COVID-19-related costs as cases decline
even more uncertainty in what hospitals’ actual and hospitals become better at managing the disease,
input price inflation will be in 2023 relative to and the continued statutory 0.5 percent increase
the forecast CMS used when setting the annual to inpatient operating payments to reverse prior
IPPS and OPPS updates for 2023, but CMS’s temporary reductions in payments that recouped prior
latest forecast suggests that the agency may have coding-induced overpayments.
underestimated 2023 input price inflation as well. The exact level of hospitals’ Medicare margin in
The underestimated inflation in 2022 and 2023 2023 will depend in large part on the duration and
contrasts with prior years: From 2012 to 2021, CMS severity of the coronavirus pandemic and associated
overestimated input price inflation in all but one PHE-related payment increases and whether the
year, for a cumulative overestimate of 5.5 percent. federal government enacts any additional coronavirus
• Expected expiration of federal relief funds and pandemic support. In addition, hospitals’ 2023
Medicare PHE payment changes, which were higher Medicare margin may be affected by CMS’s decisions
than hospitals’ additional costs. In both 2020 and on how to comply with the Supreme Court’s recent
2021, we found that hospitals’ Medicare margins ruling requiring the agency to reverse previous cuts
increased in part because the federal relief funds to OPPS payments for drugs furnished by hospitals
and Medicare payment changes during the PHE participating in the 340B drug program.31
exceeded hospitals’ additional costs from the PHE.
However, these additional payments may expire
in 2023. (The last phase of Provider Relief Fund How should Medicare payments
payments—a portion of which supports providers’
change in 2024?
care of FFS Medicare beneficiaries—began to
be distributed in early fiscal year 2022. The 2
Our payment adequacy indicators suggest that
percent sequestration of Medicare payments was
Medicare payments to general ACHs were broadly
suspended from May 1, 2020, through March 31,
adequate in 2021, and we project that they will decline
2022, and then phased in at a 1 percent reduction
in 2023 but remain broadly adequate.
through June 30, 2022, when the full 2 percent
sequestration resumed. The additional 20 percent Under current law, Medicare’s base payment rates
payment for COVID-19 inpatient stays will be under the IPPS and OPPS increase annually based on
in effect through the end of the PHE, which is the forecasted increase in the hospital market basket
currently scheduled to extend through mid-May less a forecasted increase in productivity, as well as by
2023.) any other statutory updates (see Table 3-6).
• Declines in Medicare’s uncompensated care The final updates for 2024 will not be set until summer
payments. In 2021, 2022, and 2023, Medicare’s 2023, but CMS currently forecasts a 2.9 percent
uncompensated care pool declined from $8.3 increase in the IPPS operating base payment rate and
billion to $7.2 billion to $6.9 billion, respectively. OPPS base payment rate and a 2.4 percent increase in
These declines reflect CMS’s projected drop in DSH the IPPS capital base payment rate. These forecasts,
payments and in the national uninsured rate. As based on historical data through June 2022, anticipate
Medicare payments for uncompensated care do not a marked slowdown in input price inflation. The final
80 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
TABLE
IPPS and OPPS updates and forecast errors
3–6
2019 2020 2021 2022 2023
Updates based on forecasts
IPPS operating
Market basket 2.9% 3.0% 2.4% 2.7% 4.1%
Productivity −0.8 −0.4 0.0 −0.7 −0.3
Subtotal 2.1 2.6 2.4 2.0 3.8
Statutory updates −0.3 0.5 0.5 0.5 0.5
Total 1.8 3.1 2.9 2.5 4.3
IPPS capital
Total (market basket) 1.4 1.5 1.1 1.1 2.5
OPPS
Total (same as IPPS operating subtotal) 2.1 2.6 2.4 2.0 3.8
Actual market basket
IPPS operating
Market basket 2.4 2.0 3.1 5.7* 4.3*
Forecast error 0.5 1.0 −0.7 −3.0* −0.2
IPPS capital
Market basket 1.4 1.2 1.0 1.9* 2.6*
Forecast error 0.0 0.3 0.1 −0.8* −0.1*
Note: IPPS (inpatient prospective payment systems), OPPS (outpatient prospective payment system). Updates do not include budget neutrality
adjustments to base rates. Not all IPPS and OPPS payments are increased by the updates to base rates, such as separately payable drugs.
*Data include historical data through the second quarter of calendar year 2022 and forecasts for subsequent quarters.
Source: MedPAC analysis of IPPS and OPPS final rules and data provided by CMS Office of the Actuary.
2024 update will include newer forecasts of growth in • maintain payments close to hospitals’ cost of
input prices and productivity and thus could be lower providing high-quality care efficiently;
or higher than the current projected update.
• maintain fiscal pressure on hospitals to constrain
In considering how Medicare payments to general costs;
ACHs should change in 2024, the Commission contends
that scarce Medicare resources should be used • minimize differences in payment rates for similar
efficiently. To meet this goal, Medicare should aim to services across sites of care; and
balance several objectives: • avoid implementing large, across-the-board
• maintain payments high enough to ensure payment rate increases to support a subset of
beneficiaries’ access to care; hospitals with specific needs.
Report to the Congress: Medicare Payment Policy | March 2023 81
The Commission’s payment update recommendation IMPLICATIONS 3-1
for 2024 reflects the most recent inflation and Spending
other data from 2021, preliminary data from 2022,
• We expect the recommendation to increase
and projections for 2023. If current projections of
spending relative to current law by over $2 billion
input inflation and hospital costs turn out to be
in 2024 and by over $10 billion over five years.
inaccurate, these discrepancies will be accounted for
in our assessment of payment adequacy in our next Beneficiary and provider
recommendation cycle. • The increase in Medicare payment rates will help
maintain hospitals’ willingness to treat Medicare
RECOMMENDATION 3-1 beneficiaries and maintain beneficiaries’ access to
For fiscal year 2024, the Congress should update care.
the 2023 Medicare base payment rates for general
acute care hospitals by the amount specified in
current law plus 1 percent.
Supporting Medicare safety-net
RATIONALE 3-1 hospitals
Hospitals’ payment adequacy indicators were generally The Medicare program strives to ensure access to care
positive in 2021. The number of hospital closures was for all beneficiaries and to adequately pay providers
the same as the number of openings, IPPS hospitals’ for that access. Recommendation 3-1 above is designed
all-payer operating margin increased to a record high, to provide adequate payment to the average hospital.
and IPPS hospitals’ Medicare margin increased. In other However, that level of payment may not be adequate
words, federal relief funds and increased Medicare to sustain access for Medicare beneficiaries at certain
payments more than offset pandemic-induced costs. Medicare safety-net hospitals. Therefore, Medicare’s
safety-net policies may need to be modified.
However, indicators continued to vary substantially
across hospitals, and some indicators remained below Medicare currently makes safety-net payments to
prepandemic levels. For example, some hospitals faced hospitals in the form of disproportionate share hospital
capacity and staffing constraints at times. In addition, (DSH) payments and uncompensated care payments.
FFS beneficiaries’ risk-adjusted hospital mortality rate But there are several problems with the formulas
remained higher than the rate in 2019, and patient currently used to distribute safety-net payments. First,
experience measures declined. DSH payments are applied only to hospital inpatient
rates, so hospitals get no boost to the payments they
For 2023, we project that IPPS hospitals’ Medicare
receive for providing outpatient care. Second, the
margin will decrease to about –10 percent (similar to
DSH formula is primarily driven by Medicaid patient
the level in 2017) and that relatively efficient hospitals’
shares and does not factor in Medicare patient shares.
median Medicare margin will decline to modestly
Thus, Medicare subsidizes Medicaid through its DSH
below break-even, similar to prepandemic levels.
payments and hospitals that serve high shares of
The Commission anticipates that a 2024 update to Medicare patients may be disadvantaged under the
hospital payment rates of current law plus 1 percent DSH formula. It is important for hospitals that treat
would generally be adequate to maintain FFS Medicare large shares of Medicaid patients to be supported,
beneficiaries’ access to hospital inpatient and but that cost should be Medicaid’s responsibility
outpatient care and keep IPPS and OPPS payment and not be absorbed by Medicare. Third, Medicare’s
rates close to the cost of delivering high-quality care uncompensated care payments are biased toward
efficiently. providing greater uncompensated care payments
to hospitals with few Medicare FFS inpatient stays
However, this update may not be sufficient for and more Medicare Advantage (MA) inpatient stays.
Medicare safety-net hospitals with a poor payer mix. A These issues are discussed in more detail in our June
separate discussion of how to support Medicare safety- 2022 report to the Congress and the appendix to this
net hospitals follows. chapter. A new Medicare safety-net policy should
82 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
improve how safety-net providers are identified by the number of Medicare patients and few commercial
Medicare program and the mechanisms for distributing patients, an unforeseen deviation in the profitability
Medicare safety-net payments. of Medicare patients may be far more challenging to
manage. Medicare may want to provide these safety-
Safety-net payments are warranted for net hospitals with higher payments to give them a
providers serving low-income Medicare “cushion” to account for uncertainty regarding the
beneficiaries future profitability of their Medicare patients.
We identify Medicare safety-net hospitals as those
that disproportionately serve low-income Medicare Given Medicare safety-net hospitals’ greater unpaid
patients, uninsured patients, or Medicare patients coinsurance, higher costs of low-income Medicare
that are not materially profitable. For ACHs, Medicare beneficiaries, and lack of ability to absorb unforeseen
patients—in particular, low-income Medicare patients— variation in Medicare profits, we maintain that
generate lower levels of profitability than hospitals’ supplemental payments to hospitals disproportionately
commercial patients for two reasons: serving low-income Medicare beneficiaries are
warranted. The theoretical frameworks for determining
Lower revenues per service—From 2011 to 2020, Medicare safety-net status and determining whether
IPPS hospitals’ aggregate Medicare margin has supplemental payments are necessary were discussed
been negative, ranging between −5 percent and −10 in detail in our June 2022 report to the Congress
percent, suggesting Medicare is not a profitable (Medicare Payment Advisory Commission 2022c). The
payer in aggregate in the hospital sector (Medicare Commission’s method of gauging hospitals’ safety-
Payment Advisory Commission 2022d). In addition, net status is Medicare-centric by design; safety-net
hospitals serving a high share of low-income Medicare definitions used by Medicaid and other payers likely
beneficiaries tend to receive less cost sharing because will differ from our definition.
of beneficiaries’ lack of supplemental insurance or
Medicaid not paying cost sharing for dual-eligible A new Medicare Safety-Net Index will direct
beneficiaries. Receiving less cost sharing results in safety-net payments to hospitals with high
higher levels of Medicare bad debt at Medicare safety- shares of low-income Medicare patients
net hospitals. To address the issues with the current DSH and
uncompensated care payment metrics and better
Higher costs per service—Research has indicated that direct supplemental payments to hospitals that care
hospitals’ costs per discharge for low-income Medicare for a high share of Medicare beneficiaries with low
beneficiaries are slightly higher than costs for higher- incomes, we developed a new measure called the
income beneficiaries with similar diagnoses (Nguyen Medicare Safety-Net Index (MSNI). Each hospital’s
and Sheingold 2011). MSNI is computed using three components: (1) the
The combination of lower revenue and higher costs share of its Medicare volume associated with low-
can financially strain Medicare safety-net hospitals that income beneficiaries (identified as those who receive
have to compete for labor and technology with more the Part D low-income subsidy (LIS)—see text box, p.
profitable hospitals. 84, on identifying low-income Medicare beneficiaries
and hospitals that serve them); (2) the share of revenue
In addition, hospitals that serve high shares of the hospital spends on uncompensated care (bad debts
Medicare beneficiaries and in particular high shares of and charity care); and (3) the share of total volume
low-income beneficiaries may be less able to absorb associated with Medicare beneficiaries. Table 3A-1
unforeseen financial challenges. For example, as CMS (p. 97 in the appendix to this chapter) provides more
forecasts input price inflation and then sets payment detailed information about how each hospital’s MSNI
updates accordingly, it overestimates inflation in some is calculated; the rationale for the MSNI formula is
years and underestimates inflation in other years. An discussed in our June 2022 report to the Congress
unforeseen financial challenge such as an inflation (Medicare Payment Advisory Commission 2022c). Other
forecast error is not an issue for a hospital with high payers may define safety-net status for their patients
profit margins and a large endowment. But for a differently.
hospital that just covers its expenses and has a large
Report to the Congress: Medicare Payment Policy | March 2023 83
Identifying low-income Medicare beneficiaries and the hospitals that
care for them
T
he Commission’s definition of low-income Compared with the full fee-for-service Medicare
Medicare beneficiaries includes all those who population, LIS beneficiaries are three times as
receive full or partial Medicaid benefits and likely to be disabled, nearly three times as likely
those who do not qualify for Medicaid benefits in to have end-stage renal disease, more likely to be
their states but who receive the Part D low-income female, slightly more likely to live in a rural area, and
subsidy (LIS) because they have limited assets and twice as likely to be Black or Hispanic. Given the
an income below 150 percent of the federal poverty demographic mix of the LIS population, directing
level. Collectively, we refer to this population as Medicare safety-net funds to LIS patients’ providers
“LIS beneficiaries” because Medicare beneficiaries could promote greater equity in access to care and
who receive full or partial Medicaid benefits are quality across demographic groups. More detail
automatically eligible to receive the LIS. To identify on LIS beneficiaries is provided in our June 2022
hospitals’ low-income Medicare populations, we use report to the Congress (Medicare Payment Advisory
LIS as the proxy for “low income” because it reduces Commission 2022c).
the impact of variation across states in eligibility for
Medicaid. However, the LIS definition is limited to Hospitals vary in the extent to which they care
beneficiaries who receive the LIS benefit and thus for low-income Medicare beneficiaries
omits some non-dual-eligible beneficiaries who In 2019, for the quarter of hospitals that treated the
could qualify for the LIS but have not applied for highest share of LIS beneficiaries, LIS beneficiaries
the benefit. This limitation is a result of not having made up 43 percent or more of the hospitals’
beneficiary income data. To the extent that future Medicare inpatient and outpatient volume. In
Medicare safety-net funding is attached to treating contrast, for the quarter of hospitals that treated the
more LIS beneficiaries, that payment policy would lowest share of LIS beneficiaries, LIS beneficiaries
encourage providers to make their patients aware made up 23 percent or less of the hospitals’ total
of and help them enroll in Medicaid, the Medicare Medicare volume. These data suggest that some
Savings Programs (i.e., programs in which Medicaid hospitals take on a greater responsibility for treating
helps pay for Medicare premiums, cost sharing, or low-income patients than do other hospitals, which
both), and the Part D LIS.32 could be financially disadvantageous. ■
The MSNI model and the current uncompensated care under current policy, MA safety-net payments are not
policy differ importantly in that the MSNI payments proportional to uncompensated care costs (see Table
would be structured as add-on payments to Medicare 3A-3, p. 99, in the appendix for details).
payment rates (meaning a percentage increase to FFS
rates for each claim). Providers with more financially A second difference is that DSH payments currently
challenging patient mixes would receive higher increase as the share of patients insured primarily
Medicare payment rates. In contrast, the current by Medicaid increases. Thus, Medicare subsidizes
uncompensated care model is not directly tied to Medicaid through DSH payments. Under the MSNI
Medicare payment rates. Each hospital receives a model, Medicaid patients (other than dual-eligible
fixed share of its uncompensated care costs from FFS beneficiaries) will not directly affect Medicare
Medicare. That in turn sets the add-on amount per FFS payments. While the MSNI does not directly support
claim that is used by MA plans.33 The net result is that, Medicaid, notably, hospitals with high shares of low-
84 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
income Medicare patients will benefit, and those report, the MSNI also has the benefit of not directly
hospitals typically also have high shares of Medicaid subsidizing Medicaid and not being inversely correlated
patients. with Medicare shares.
In our June 2022 report to the Congress, we used 2016 Redistributing DSH and uncompensated care
data to simulate how Medicare payments would have funds using the MSNI would increase high-MSNI
changed if the MSNI was used to distribute safety- hospitals’ Medicare revenue by about 3.9 percent
net dollars. We used 2016 data because we wanted to in aggregate
examine hospitals that closed between 2016 and 2020 We simulated how Medicare and all-payer payments
to determine the extent to which they would have been would have changed for each IPPS hospital if
helped if safety-net payments had been distributed by we redistributed the $11.7 billion of DSH and
the MSNI. We found that the MSNI would have directed uncompensated care funds that hospitals received
more dollars toward hospitals with lower all-payer in 2019 using the MSNI. To allocate the MSNI dollars
margins and to hospitals that closed from 2016 to 2020. among hospitals, we used a linear model where the
percentage add-on (to inpatient and most outpatient
In this chapter, we update our analysis to simulate
rates) increases as the MSNI increases. In this
what would have happened in 2019 if the MSNI had
illustrative example, the MSNI add-on starts at zero
been used to distribute safety-net payments rather
for hospitals with an MSNI at the 10th percentile or
than the DSH and uncompensated care policies that
below. These hospitals receive no Medicare safety-
were in effect in 2019. Like the results using 2016 data,
net payments.35 For hospitals above this threshold,
the simulation using 2019 data suggests that the MSNI
the percentage adjustment of the Medicare safety-net
would have helped redirect funds toward hospitals that
add-on continuously increases according to a linear
tended to serve lower-income Medicare beneficiaries
model. It rises to 3 percent at the 25th percentile of the
and had relatively low 2019 all-payer margins.
MSNI distribution, 8 percent at the 50th percentile, 14
The MSNI is a better indicator of financial status percent at the 75th percentile, and 26 percent at the
of hospitals serving large shares of low-income 95th percentile of the MSNI distribution. The maximum
Medicare beneficiaries than the DSH metric MSNI redistribution add-on was set at 26 percent (the
95th percentile) to avoid extreme add-ons for outlier
To compare how well the DSH metric and the MSNI
hospitals.
identify hospitals under financial strain, we examined
characteristics of hospitals that were divided into As noted above, unlike the current DSH and
quartiles based on the DSH and MSNI scores.34 The uncompensated care payments, the MSNI payment
DSH metric and the MSNI are moderately correlated add-ons would apply to both inpatient and most
(correlation = 0.56). They both have some ability to outpatient services. The one exception is separately
identify hospitals under financial strain. For example, payable Part B drugs. The acquisition costs of drugs are
hospitals in the highest quartile of both the DSH metric unlikely to be higher for Medicare safety-net hospitals,
and the MSNI tend to have greater uncompensated and they are lower than average if those hospitals
care costs, larger amounts of unpaid Medicare cost qualify for 340B status. Therefore, we excluded
sharing (Medicare bad debts), and lower all-payer total separately payable Part B drug claims from eligibility
margins. However, the MSNI appears to do a better job for the MSNI add-on to prevent an unlevel playing field
differentiating hospitals according to their level of all- where certain safety-net providers could specialize
payer profitability and financial stress. For example, in in providing expensive Part B drugs. The exclusion of
2019, the hospitals in the lowest DSH quartile had an separately payable Part B drugs is a new refinement in
aggregate all-payer total margin that was 5 percentage our method that occurred after the June 2022 report to
points higher than hospitals in the highest DSH quartile the Congress was published.
(10.1 percent vs. 5.1 percent). In contrast, hospitals in
the lowest MSNI quartile had an aggregate all-payer Our simulation allows almost all hospitals to receive
total margin that was 6.9 percentage points higher than MSNI payments. The simulation used a graduated
hospitals in the highest MSNI quartile (10.0 percent linear increase in the MSNI percentage add-on for
vs. 3.1 percent). As we discussed in our June 2022 two reasons. First, as we explained in the June 2022
Report to the Congress: Medicare Payment Policy | March 2023 85
TABLE
Hospital financial characteristics under current DSH and uncompensated
3– 7 care policy and simulations of redistributing based on the MSNI, 2019
DSH quartiles Medicare Safety-Net Index (MSNI) quartiles
Characteristic Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
Actual 2019 financial data given current DSH and uncompensated care policy
DSH payments / 0.5% 1.3% 2.0% 3.4% 1.3% 1.7% 2.1% 3.3% Current-law estimates
FFS Medicare revenue are the same as in the
Uncompensated 1.9 3.6 4.0 6.5 2.7 3.6 4.5 8.6 center block of data
care payments / FFS
Medicare revenue
Unpaid FFS Medicare 0.5 0.6 0.7 0.8 0.3 0.5 0.7 1.0
bad debts / FFS
Medicare revenue
FFS Medicare margin −11.0 −7.0 −8.9 −7.2 −12.4 −9.5 −5.5 −0.9
All-payer total margin 10.1 8.1 8.8 5.1 10.0 8.3 6.0 3.1
Illustrative simulations of distributing DSH and uncompensated care pools using MSNI payments
Adding $1 billion to
the FFS pool plus
Redistribution of existing dollars proportionate MA funding
Mean FFS Medicare $0.4 0.2 −0.1 −0.5 −$2.8 −0.3 1.5 1.5 −$2.7 0.1 2.0 2.1
payment change
(millions)
Simulated MSNI 3.1% 5.1 6.1 9.2 1.0% 5.0 8.9 15.9 1.1 % 5.4 9.6 17.2
payments / FFS
Medicare revenue
Percent change in 0.7 0.3 −0.1 −0.7 −3.0 −0.3 2.4 3.9 −2.9 0.1 3.1 5.3
FFS Medicare revenue
Percent change in total 0.3 0.1 −0.1 −0.3 −0.9 −0.1 0.8 1.0 −0.9 0.0 1.0 1.4
revenue*
Simulated FFS Medicare −10.1 −6.7 −9.0 −7.9 −15.8 −9.8 −3.1 3.0 −15.7 −9.4 −2.3 4.2
margin under MSNI
Simulated all-payer 10.4 8.2 8.8 4.8 9.2 8.2 6.7 4.0 9.2 8.3 6.9 4.4
margin under MSNI*
Note: DSH (disproportionate share hospital), MSNI (Medicare Safety-Net Index), Q (quartile), FFS (fee-for-service) , MA (Medicare Advantage). The DSH
quartiles were based on the disproportionate share patient percentage. The unit of analysis is the quartile, with payments and costs of 651 hospitals
in each quartile added to create aggregate payment changes and aggregate margins for the quartile. The aggregate margin is equivalent to a
dollar-weighted margin for the hospitals. The full sample of 2,604 hospitals represents all hospitals paid through the acute inpatient prospective
payment systems (excluding territories) with more than 200 Medicare discharges and complete data. The margins presented are the aggregate
margins for each group. This analysis differs from the analysis in the June 2022 report in that it examines 2019 data (in which current-law payments
were different) and does not allow any add-ons to Part B drug spending. In addition, we started payments at the 10th percentile of the MSNI
distribution rather than the 5th percentile to increase targeting of the payments. We also did not allow any change in payments for hospitals
currently choosing Medicare-dependent hospital status or sole community hospital status (and that thus receive payments partially based on
historical costs). Despite these changes, the results are directionally consistent with the earlier analysis of how the MSNI would have affected
payments in 2016.
*Estimates of change in aggregate all-payer margins assume that changes in Medicare Advantage (MA) payment rates equal changes in FFS
Medicare rates (i.e., that MA plans pay FFS rates) and that the ratio of MA inpatient and outpatient volume to FFS volume can be approximated by
the ratio of MA discharges to FFS discharges.
Source: MedPAC analysis of claims, cost report, and closure data.
report, we wanted to avoid a cliff where hospitals just payment. Second, about 80 percent of IPPS hospitals
below a threshold received no add-on and hospitals now receive DSH and uncompensated care payments,
above that threshold received a dramatically higher and we wanted to limit the share of these hospitals that
86 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
would not receive any MSNI payments. In the current (from 3.1 percent to 4.4 percent) (Table 3-7). In 2019, the
simulation, about 7 percent of hospitals would lose total cost of increasing FFS MSNI payments by $1 billion
their DSH and uncompensated care payments and not and a commensurate add-on for hospitals treating MA
receive any of the new MSNI payments. In 2019, these patients would have been about $1.5 billion. We expect
hospitals had an average all-payer total margin of 11.8 that an equivalent percentage add-on to FFS and MA
percent. At the same time, about 5 percent of hospitals payments in 2024 would cost close to $2 billion due to
currently do not receive any DSH or uncompensated updates in Medicare payment rates and increases in
care payments and would gain MSNI payments. In total Medicare enrollment (primarily in MA) from 2019
2019, the hospitals that would gain Medicare safety-net to 2024.
funding had an average all-payer total margin of 4.5
percent. The MSNI tends to benefit hospitals with high
Medicare shares and reduce payments to
We estimate that using the MSNI to redistribute hospitals with low Medicare shares and high
existing DSH and uncompensated care funds would uncompensated care costs
have increased Medicare payments to hospitals in Shifting safety-net payments from the current DSH
the high-MSNI quartile by 3.9 percentage points, and uncompensated care payments to new MSNI-
increasing the aggregate FFS Medicare margin from based payments would change the distribution of
−0.9 percent to 3.0 percent (Table 3-7). payments in three important ways. First, because
hospitals’ dependence on Medicare patients is a factor
In turn, the higher Medicare margin for high-MSNI
in computing the MSNI, hospitals with higher shares
hospitals would result in a smaller difference in the
of Medicare patients would tend to receive higher
all-payer total margin between hospitals in the highest
add-on payments per case. These hospitals also
and lowest MSNI quartiles. The difference would fall
would receive the MSNI add-on payment for a greater
from 6.9 percentage points (10.0 percent to 3.1 percent)
share of the services they furnish because Medicare
under current law to 5.2 percentage points (9.2 percent
is a large share of their patient mix. Second, because
to 4.0 percent) under the MSNI redistribution.
Medicare would no longer directly subsidize Medicaid
To provide an illustrative example of how changing patients, hospitals with few Medicare patients and
the pool of dollars would change the add-on payment, large Medicaid patient loads would see a reduction
we estimated the effect on hospitals if the size of the in payments. Third, because Medicare would provide
FFS DSH and uncompensated care pool of dollars only modest indirect support for uncompensated care,
were increased from the approximately $11.7 billion hospitals with low Medicare volume but high levels
that was disbursed in 2019 to an illustrative $12.7 of uncompensated care would tend to receive less
billion.36 For every billion dollars added to the MSNI funding. Under the scenario in which current DSH and
pool, overall Medicare FFS hospital spending would uncompensated care dollars would be redistributed
increase by about a half percent. The net effect is and $1 billion would be added to the FFS pool of safety-
that the add-on would grow from zero at the 10th net funds, hospitals would still experience a decline in
percentile of the distribution to about 29 percent at revenue if their current uncompensated care payments
the 95th percentile and above in the distribution. In from Medicare were larger than the value of the
2019, MSNI hospitals in the top quartile (which would proposed MSNI add-on payments (up to 29 percent).
receive a disproportionate share of any additions to Overall, payments would shift toward hospitals serving
the MSNI pool) would have seen their FFS Medicare high volumes of Medicare patients and, in particular,
margin increase by about 5.1 percentage points (from low-income Medicare patients.
−0.9 percent to 4.2 percent). Our simulation assumed
To provide a greater understanding of which types
that CMS would provide hospitals serving MA patients
of IPPS hospitals would gain and lose under a shift
with a commensurate adjustment; those additional MA
from the current DSH and uncompensated care
payments would have totaled about $0.5 billion in 2019.
payments to payments based on the MSNI, we
The combination of additional FFS and MA payments
provide some descriptive statistics on payment
would cause high-MSNI hospitals’ all-payer total
changes for 10 categories of hospitals under our MSNI
margins to increase by about 1.3 percentage points
Report to the Congress: Medicare Payment Policy | March 2023 87
TABLE
Simulated effect of redistributing current DSH and UC payments under the
3–8 MSNI and adding $1 billion to the FFS MSNI pool by type of hospital, 2019
Percentile effect on
Aggregate all-payer total margins
percentage change in: in percentage points**
Mean MSNI
percentage
Hospital add-on to FFS FFS Medicare All-payer
characteristic Medicare payments* revenue total revenue 5th 25th 75th 95th
All IPPS hospitals 10.4% 0.5% 0.1% −1.7% −0.4% 1.7% 4.6%
Government (n = 349) 14.0 −1.5 −0.6 −2.6 −0.5 1.6 4.6
For profit (n = 592) 11.6 2.3 0.8 −1.7 0.0 2.4 5.7
Nonprofit (n = 1,663) 9.2 0.5 0.2 −1.6 −0.4 1.4 4.2
Rural (n = 611) 13.7 3.3 1.1 −0.4 0.0 2.6 5.7
Urban (n = 1,990) 9.3 0.2 0.1 −2.0 −0.6 1.4 4.2
Teaching (n = 1,568) 10.1 0.1 0.0 −2.3 −0.6 1.4 4.0
Nonteaching (n = 1,033) 10.5 1.3 0.4 −1.4 0.0 1.8 5.1
MA share of stays
< 25% (n = 1,308) 9.7 0.7 0.2 −1.2 0.0 1.4 5.0
25% to 50% (n = 949) 10.2 0.5 0.2 −1.7 −0.4 1.9 4.3
> 50% (n = 347) 12.5 −0.3 −0.1 −3.1 −1.0 1.6 4.4
Note: DSH (disproportionate share hospital), UC (uncompensated care), MSNI (Medicare Safety-Net Index), FFS (fee-for-service), IPPS (inpatient
prospective payment systems), MA (Medicare Advantage). The table presents unweighted mean values comparing payments that occurred in
2019 with what payments would have been under an MSNI distribution of safety-net dollars. Data include all IPPS hospitals in the United States
(excluding territories) with more than 200 discharges and complete cost report data in 2019. The 5th and 95th percentiles on the right-hand side of
the table illustrate that 5 percent had a reduction equal to or larger than the 5th percentile and 5 percent had an increase equal to or larger than
the 95th percentile in our 2019 simulation.
*Add-on adjustments are applied to inpatient and outpatient payments excluding Part B drugs.
**Estimates of change in total margins assume that MA plans shift payment rates to equal the shift in FFS payment rates and that the ratio of MA to
FFS volume can be estimated using the ratio of MA discharges to FFS discharges.
Source: MedPAC analysis of cost report and claims data.
model and the addition of $1 billion to the FFS MSNI respectively) (Table 3-8). Nevertheless, because
pool (Table 3-8).37 Most of the hospitals that would some government hospitals currently receive high
gain under the redistribution are smaller hospitals levels of DSH and uncompensated care payments
with higher Medicare shares. Most hospitals that relative to their Medicare volume, under the MSNI,
would experience a reduction in payments are larger government hospitals in aggregate would see a
hospitals that currently receive high uncompensated decline in FFS Medicare payments of 1.5 percent
care payments. (Table 3-8).38 Rural hospitals in aggregate would see
an increase in FFS Medicare payments of 3.3 percent.
On average, government-owned hospitals and Such hospitals would benefit because they tend to
rural hospitals would receive the highest MSNI have high Medicare shares that are not factored into
add-on percentages (14.0 percent and 13.7 percent,
88 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
current DSH percentages or uncompensated care MSNI pool of funds. About 20 percent of hospitals
payments. Rural hospitals would also tend to benefit would increase their total revenue by more than
from removing the distortion in uncompensated care 2 percent, with the largest gains often going to
payments that direct payments to hospitals with high hospitals with high FFS Medicare shares and few MA
MA shares (see Table 3A-3, p. 99, in the appendix to patients. In contrast, about 4 percent of hospitals
this chapter). In addition, for-profit hospitals would would have their total revenue reduced by at least
be slightly more likely to benefit from the policy than 2 percent. These hospitals tend to have relatively
nonprofit hospitals because the Medicare patients high uncompensated care payments and relatively
served by for-profit hospitals are more likely to small shares of FFS Medicare patients. Their revenue
receive the LIS. In 2019, for-profit hospitals had an would see a decrease as Medicare safety-net
LIS share of 36 percent compared with nonprofits’ payments transitioned away from directly funding
32 percent share (data not shown). (Government uncompensated care to focus on assisting hospitals
hospitals had an LIS share of 42 percent.) with high Medicare shares and, in particular, high
shares of low-income Medicare patients.
Shifting to MSNI-based payments would also tend to
increase add-on payments for hospitals with low MA Because there are some hospitals that would face
shares of Medicare admissions and decrease payments material reductions in revenue under the MSNI policy,
for hospitals with higher shares of MA admissions the Congress could phase in the policy for all hospitals
(Table 3-8). This result could be due to several factors. over a set period of time (e.g., transition to the
One factor is that the current method of distributing MSNI policy over three to five years). Alternatively, a
uncompensated care funds favors hospitals with few transition could be managed through a stop-loss policy
FFS Medicare stays and higher numbers of MA stays so that no hospital would experience changes (positive
(see Table 3A-3, p. 99, in the appendix to this chapter). or negative) in Medicare payments of more than 5
The new MSNI method would address this issue.39 percent in any one year because of the transition. This
change would produce a variable transition, with some
In Table 3-8, we present the 5th percentile of hospitals, hospitals fully transitioned to the MSNI payments
which would see the largest declines in total margins, sooner than others. Both approaches would allow time
and the 95th percentile of hospitals, which would see for the hospitals facing the most substantial revenue
the largest increases in total margins. The similarity reductions to try to augment revenues from existing
across hospital types in total margin changes at the sources and request additional financial support from
5th and 95th percentiles indicates that the distribution state and local governments, as warranted. The portion
of changes in Medicare revenue would be similar of these hospitals with high cost structures may also be
across the different types of hospitals. Under the MSNI able to improve efficiencies.
policy, we expect that about 5 percent of providers in
all categories would experience declines of at least 1 Incorporating the MSNI across FFS and MA
percent to 2 percent of all-payer total revenue, while
Nearly half of Medicare beneficiaries are enrolled
about 5 percent of providers in all categories would
in MA.40 Therefore, the MSNI was calibrated using
experience at least a 4 percent to 5 percent increase
FFS and MA data (when possible) and should be
in revenue. The percentage increases at the top end of
applied to hospital care provided to both MA and FFS
the tail are larger than the decreases at the bottom end
beneficiaries. MA beneficiaries should be included
of the tail because smaller hospitals tend to gain more
when computing the MSNI because MA plans largely
with the MSNI. In other words, a 1 percent to 2 percent
pay hospitals rates similar to FFS Medicare and
decline in payments to a larger hospital can fund a 3
hospitals likely incur similar costs for treating MA and
percent to 4 percent increase in payments to smaller
FFS beneficiaries.41
hospitals.
Policymakers may choose different ways to incorporate
Most hospitals in our simulation saw increases
the FFS MSNI payments (which are structured as
in Medicare payment rates under the MSNI
claims-based add-ons) into MA per capita payments.
policy because of the additional $1 billion in FFS
Under a preferred pathway, CMS would calculate an
uncompensated care payments added to the
Report to the Congress: Medicare Payment Policy | March 2023 89
MSNI add-on percentage for each hospital, calculate (see Table 3A-3, p. 99, in the appendix to this
MSNI add-on payments by applying the add-on chapter for an example).
percentage to each hospital’s encounter claims for
MA beneficiaries, and directly provide the add-on Interactions between the MSNI and other
payment to hospitals, bypassing the MA plan itself Medicare special payments
in the transaction.42 Because the MSNI funds would The MSNI is intended to compensate hospitals for
be paid directly to hospitals, MSNI payments for FFS the higher costs and lower revenues associated
beneficiaries would be excluded from MA benchmarks, with treating a high share of Medicare beneficiaries,
comparable with the way indirect medical education particularly low-income Medicare beneficiaries, and
payments are currently made to hospitals for their FFS patients without insurance. However, many hospitals
and MA patients. We believe there are several benefits already receive special payment rates from Medicare
to this approach: that help compensate for these costs (to the extent
that hospitals incur them) and help maintain access
• Safety-net payments would flow directly to
in certain areas or for specific populations. Such
Medicare safety-net providers and would not
hospitals include sole community hospitals, Medicare-
simply represent additional funds for MA plans to
dependent hospitals, critical access hospitals, and
use at their discretion (which might be the case if
rural community hospital demonstration hospitals.
MSNI payments were included in benchmarks).
Policymakers should consider requiring hospitals
• MA plans would not have an incentive to exclude to choose between retaining their current special
safety-net providers with high MSNI add-on payment designations or receiving the IPPS rates with
payments from their networks. Under current an MSNI supplement. Giving hospitals the option to
regulations, MA plans have an incentive to exclude choose their preferred payment mechanism maximizes
hospitals with high DSH and uncompensated care flexibility for hospitals while making sure Medicare
add-ons from their networks because MA plans payments are not excessive (e.g., providing a 20 percent
often pay FFS rates. MSNI add-on to a hospital that Medicare already
pays on a cost basis). In addition, policymakers may
• Linking funds to encounter data could incentivize want to cap interactions with other existing IPPS
providers to encourage MA plans to improve their payment adjustments, such as the low-volume hospital
submission of encounter data. adjustment (which increases IPPS payments by up to 25
percent), such that the maximum cumulative add-on to
• A key distortion in the way uncompensated
IPPS payments from all special designations could not
care payments affect MA benchmarks would be
exceed a specified threshold (e.g., 30 percent).
removed. Currently, Medicare’s uncompensated
care add-on payment per FFS discharge varies
Recommendation
such that the aggregate additional FFS add-on
payments to each hospital equals a common To better target hospitals serving low-income
expected percentage (e.g., 20 percent) of all Medicare beneficiaries and fix adverse incentives with
hospitals’ historical uncompensated care costs. the current DSH and uncompensated care payments,
This add-on can increase FFS payment rates by current safety-net payments should be redirected to
30 percent or more (even 100 percent at hospitals hospitals that have high MSNIs. These MSNI-based
with few FFS discharges). These higher FFS payments would adjust for the lower cost sharing
payments are incorporated into MA benchmarks. received by high MSNI hospitals, the higher costs of
Because the uncompensated care payments are low-income Medicare patients, and the need for an
spread only across FFS discharges, the current additional support to absorb any unforeseen costs of
policy favors hospitals with higher MA penetration serving Medicare beneficiaries. Without additional
(that is, with fewer FFS discharges over which safety-net funding, unforeseen reductions in the
to spread the additional payments). Fewer FFS profitability of serving Medicare beneficiaries could
discharges results in a higher adjustment per FFS be difficult to absorb for hospitals with high Medicare
discharge and a larger increase in MA benchmarks shares and few commercial patients.
90 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
RECOMMENDATION 3-2 Programs (i.e., programs in which Medicaid helps pay
In fiscal year 2024, the Congress should: for Medicare premiums and cost sharing), and the
Part D LIS.
• begin a transition to redistribute
disproportionate share hospital and CMS should use a graduated linear increase in the
uncompensated care payments through the MSNI percentage add-on amount. Using a continuous
Medicare Safety-Net Index (MSNI); scale will minimize payment “cliffs” and ensure that
• add $2 billion to the MSNI pool; most hospitals continue to receive some level of
Medicare safety-net funding.
• scale fee-for-service MSNI payments in
proportion to each hospital’s MSNI and MSNI payments should be structured as add-on
distribute the funds through a percentage payments to Medicare payment rates and should
add-on to payments under the inpatient and
apply to services provided under both the IPPS and
outpatient prospective payment systems; and
the OPPS. However, coinsurance would continue to be
• pay commensurate MSNI amounts for services based on the pre-MSNI payment amount to ensure that
furnished to Medicare Advantage (MA) enrollees beneficiaries using Medicare safety-net hospitals would
directly to hospitals and exclude them from MA not pay more than patients at hospitals with fewer low-
benchmarks.
income patients. However, separately payable Part B
Each hospital’s MSNI should be computed using three drugs should not be eligible for MSNI add-on payments
components: (1) the share of its Medicare volume because the acquisition costs of these drugs are
associated with low-income beneficiaries, (2) the share unlikely to be higher for Medicare safety-net hospitals
of its revenue the hospital spends on uncompensated (and may even be lower for hospitals that qualify for
care (bad debts and charity care), and (3) the share of 340B status). Further, including separately payable
total volume associated with Medicare beneficiaries. Part B drugs in this policy would create incentives for
Under this computation, Medicare would no longer certain safety-net providers to specialize in providing
subsidize Medicaid, as it does through its DSH expensive Part B drugs.
payments. While the Commission emphasizes that
Policymakers should consider calculating MSNI add-on
scarce Medicare funds should be used to support care
payments for services provided to MA enrollees by
provided to Medicare beneficiaries and not cross-
applying the add-on percentage to each hospital’s
subsidize the care of other patient populations, we
encounter claims for MA beneficiaries and paying
also acknowledge that replacing the current DSH and
the resulting amount directly to hospitals. In doing
uncompensated care payments with new MSNI-based
so, MSNI payments for FFS beneficiaries would be
payments would lead to material changes for certain
excluded from MA benchmarks. This method would be
hospitals. A multiyear transition would provide time
similar to the way indirect medical education payments
for hospitals and other payers to adjust to this new
are currently made to hospitals for their FFS and MA
Medicare payment approach. (See text box, pp. 92–93,
patients. Making MSNI payments for MA enrollees
on the future roles of Medicare and Medicaid payments
directly to hospitals would reduce current incentives
in supporting Medicare and Medicaid safety-net
for MA plans to steer patients away from Medicare
hospitals.)
safety-net hospitals with high MSNI add-on payments.
CMS should define low-income Medicare beneficiaries At the same time, linking funds to encounter data could
as those who receive full or partial Medicaid benefits incentivize hospitals to encourage MA plans to improve
and those who do not qualify for Medicaid benefits in their submission of encounter data.
their states but who receive the Part D LIS because
Many hospitals already receive special payment rates
they have limited assets and an income below 150
from Medicare to help ensure access to care for
percent of the federal poverty level. Using this
Medicare beneficiaries, including sole community
definition would reduce the impact of variation in state
hospitals, Medicare-dependent hospitals, critical
Medicaid policies on Medicare payment and could
access hospitals, and rural community hospital
encourage providers to make their patients aware of
demonstration hospitals. Policymakers should
and help them enroll in Medicaid, the Medicare Savings
Report to the Congress: Medicare Payment Policy | March 2023 91
The future roles of Medicare payments and Medicaid payments in supporting
Medicare safety-net hospitals and Medicaid safety-net hospitals
T
o ensure access to care for patients with low There are three situations under which Medicare
incomes, it is important to understand the and Medicaid can make DSH or uncompensated
roles of Medicaid and Medicare in supporting care payments to help cover the costs of the same
Medicaid patients, Medicare dual-eligible patients, uninsured patients. First, uncompensated care costs
and uninsured patients. In 2019, the Medicaid for the uninsured can both be partially covered
program made about $13 billion of Medicaid by the Medicare uncompensated care pool and be
disproportionate share hospital (DSH) payments a qualified expense for Medicaid DSH payments.
(which differ from Medicare DSH payments) to When Medicaid computes uncompensated care
acute care hospitals (Medicaid and CHIP Payment costs, the program measures the amount of
and Access Commission 2022a). States have broad uncompensated care provided to uninsured patients.
discretion in how their Medicaid DSH and other These costs are computed prior to any payments
safety-net funds are allocated among hospitals. In by Medicare from its uncompensated care pool and
some states, Medicaid DSH payments are highly are therefore measured as gross uncompensated
concentrated at a few safety-net hospitals, while care costs and not costs net of Medicare support.
in other states they are more widely distributed Second, some states have also received Section
across almost all hospitals. States can choose 1115 demonstration authority to make Medicaid
whether to allocate these funds based on hospitals’ uncompensated care pool payments that are
uncompensated care burdens. However, states similar to DSH payments and may similarly pay
must limit each individual hospital’s Medicaid DSH for uncompensated care. Third, states can require
payments to the sum of that hospital’s Medicaid Medicaid managed care plans to provide “directed
shortfall (the difference between the hospital’s payments” to specific hospitals. For example, the
Medicaid costs and Medicaid revenues) and state can mandate higher Medicaid rates to safety-
the uncompensated care costs associated with net hospitals (Medicaid and CHIP Payment and
uninsured patients.43 Access Commission 2022b).
(continued next page)
consider requiring such hospitals to choose between RATIONALE 3-2
retaining their current special payment designations or
The Commission’s analyses have shown, on average,
receiving IPPS rates with an MSNI supplement. Giving
that Medicare beneficiaries have good access to
hospitals the option to choose their preferred payment
hospital care, and hospitals’ total (all-payer) margins
mechanism maximizes flexibility for hospitals while
are near record highs as a result of rapidly increasing
ensuring that Medicare payments are not excessive. In
rates paid by commercial insurers. However, hospitals’
addition, policymakers may want to cap interactions
aggregate Medicare margin is negative and near
with other existing IPPS payment adjustments (e.g.,
zero even for relatively efficient hospitals. For ACHs,
the low-volume hospital adjustment) such that the
Medicare patients—in particular low-income Medicare
maximum cumulative add-on to IPPS payments from
patients—generate lower revenues per service and may
all special designations could not exceed a specified
be associated with higher costs. This lower level of
threshold (e.g., 30 percent).
profitability may render Medicare safety-net hospitals
92 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
The future roles of Medicare payments and Medicaid payments in supporting
Medicare safety-net hospitals and Medicaid safety-net hospitals (cont.)
In recent years, Medicaid has shifted away from patients) and would continue to directly support the
supporting Medicare dual-eligible patients. Under uninsured through Medicaid DSH payments.
the Consolidated Appropriations Act, 2021 (P.L. 116-
260), beginning in October 2021, losses on Medicare If Medicare shifted to providing safety-net
dual-eligible beneficiaries can no longer be used support based on the MSNI, direct support for
to justify Medicaid DSH payments.44 Medicaid uncompensated care would come only from
DSH payments now focus almost exclusively on Medicaid, and Medicaid’s uncompensated care
the costs of Medicaid patients and the uninsured. support would be limited to covering the costs
Consistent with this approach, the Commission’s of the uninsured. The Medicare program would
recommendation that Medicare begin to redistribute indirectly support uncompensated care, but most
DSH and uncompensated care payments through Medicare safety-net funding would be focused
the Medicare Safety-Net Index (MSNI) would on supporting the costs of low-income Medicare
distinguish Medicare’s responsibilities from those of patients. The Commission takes no position on
Medicaid. Under the MSNI proposal, the Medicare whether hospitals with high Medicaid patient shares
program would have full responsibility for Medicare and uncompensated care burdens should receive
patients (including dual-eligible patients) but would more or less funding from Medicaid and local
no longer provide higher Medicare payments for governments. However, the Commission asserts
hospitals with greater Medicaid patient shares. that, just as Medicaid DSH payments are focused
The recommendation would also limit Medicare’s on hospitals with high uninsured and Medicaid
support of uncompensated care and tie that support shares, Medicare safety-net payments should be
to a hospital’s Medicare volume. By contrast, the used primarily to support hospitals that provide care
Medicaid program would have full responsibility to larger shares of low-income Medicare patients
for Medicaid patients (excluding dual-eligible rather than the uninsured or Medicaid patients. ■
less able to absorb unforeseen financial challenges and The MSNI would not only create greater financial
can undermine their ability to compete with wealthier stability for hospitals serving high shares of low-
hospitals for labor and technology. income Medicare patients, it would also increase all
hospitals’ incentives to serve Medicare patients. The
Medicare already provides substantial safety-net Commission estimates that the marginal profit on
funding to hospitals, but there are several problems Medicare patients was about 8 percent in 2021. This
with the way Medicare distributes these funds, marginal profit is computed by examining marginal
including omitting a hospital’s Medicare share from revenue from serving Medicare patients compared with
its funding formulas in favor of subsidizing Medicaid marginal costs. Because FFS uncompensated care (UC)
payments and making supplemental payments only for payments are not tied to Medicare volume, they do not
inpatient services. The Commission-developed MSNI increase as the hospital serves more Medicare patients.
better identifies hospitals at financial risk and would Thus, UC payments are not part of marginal revenue
better focus scarce Medicare resources to support per unit of service. In contrast, if Medicare’s UC
hospitals that are key sources of care for low-income payments were distributed via the MSNI, they would
Medicare beneficiaries. become add-on payments and increase the marginal
revenue of serving Medicare patients. In aggregate,
Report to the Congress: Medicare Payment Policy | March 2023 93
hospitals’ marginal profit on providing inpatient and funds, and the share of MA benchmarks derived from
outpatient services to FFS Medicare beneficiaries that spending. With this transition, the MSNI would
would increase from about 8 percent to about 12 remove the adverse incentives and inaccurate targeting
percent.45 of Medicare’s current safety-net payments. However,
DSH computations could still be made to determine
The Commission anticipates that adding a combined eligibility for certain programs, such as the 340B
$2 billion in additional Medicare payments for FFS and program.
MA payments would be enough to help maintain the
financial viability of Medicare safety-net hospitals. (The IMPLICATIONS 3-2
magnitude of the $2 billion add-on could grow annually
by the hospital market basket.) Spending
• We expect the recommendation to increase
We expect the additional funds to immediately be spending relative to current law by between $750
distributed through the MSNI in 2024 and future years, million and $2 billion in 2024 and by over $10 billion
thus increasing MSNI hospitals’ FFS and MA payments over five years.
by about $2 billion in each year. In addition, over a
Beneficiary and provider
period of years, the current FFS DSH funds, FFS UC
funds, and the share of MA benchmarks derived from • We expect the recommendation to increase
those funds would be transferred to the MSNI pool. hospitals’ willingness and ability to treat low-
Eventually the MSNI would replace all DSH funds, UC income Medicare beneficiaries. ■
94 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
3-AA P P E N D I X
Supplemental
information on the
Medicare Safety-Net Index
Developing the Medicare Safety-Net Index In IPPS rulemaking for fiscal year 2014, CMS originally
Table 3A-1 explains in detail how the MSNI was created. proposed to make interim UC payments on a periodic
It is designed to target financially vulnerable hospitals basis (not associated with any claims). However, in
and in particular hospitals that serve large shares of the final rule, CMS implemented a policy that makes
low-income Medicare beneficiaries. These hospitals interim UC payments as add-on payments to each
are vulnerable to unforeseen circumstances (such as FFS IPPS discharge.46 CMS made this change after
the underestimate of input price inflation) that cause hospitals raised concerns that almost all hospitals
Medicare program payments to be lower than optimal had contracted with MA organizations to use the
in some years and too high in other years. Medicare IPPS Pricer software in setting their payment
rates, resulting in MA plans generally paying FFS rates
The new Medicare safety-net payment (Berenson et al. 2015, Centers for Medicare & Medicaid
would continue the evolution of Medicare’s Services 2013, Maeda and Nelson 2017). Therefore,
hospital safety-net payments away from if CMS had excluded UC payments from the IPPS
traditional DSH payments payment rate, MA plans would have reduced payments
To put the proposed MSNI change to safety-net to DSH hospitals (Centers for Medicare & Medicaid
payments in context, Table 3A-2 (p. 98) provides Services 2013). Because CMS decided to include the
a summary of how Medicare safety-net payments UC payments in the IPPS Pricer, MA plans make UC
have evolved over time. The table explains how the payments to hospitals on a per discharge basis similar
original DSH payments supported hospitals with to FFS UC payments per discharge.
higher Medicaid shares and higher Medicare inpatient
For 2023, CMS established the FFS UC pool at $6.9
volume, how the program shifted in 2018 away from
billion dollars. Since DSH hospitals’ historical UC costs
supporting higher Medicaid and Medicare volumes to
were about $34 billion, each DSH hospital therefore
more directly supporting uncompensated care, and
will receive approximately 20 percent of its historical
how the MSNI would shift support to hospitals with
UC costs as FFS add-on payments. These add-on
higher shares of low-income Medicare patients. Should
payments are distributed as an interim payment per
the MSNI approach be adopted, the program will have
FFS discharge based on the hospital’s historical FFS
evolved from directing higher payments to providers
stays.47 In addition, because MA plans generally pay
with high Medicaid shares, then to hospitals with high
the same rates as those under FFS, hospitals effectively
uncompensated care costs, and finally, under the MSNI,
receive UC payments from MA plans for their enrollees
to supporting hospitals that serve a disproportionate
on a per discharge basis, similar to FFS UC payments
share of low-income Medicare beneficiaries.
per discharge.
Medicare’s current method of distributing However, as illustrated in Table 3A-3 (p. 99), the current
UC payments is biased against hospitals system is inequitable. Because the FFS interim per
with high FFS volume and low MA volume discharge UC payments are included in the IPPS Pricer
The Affordable Care Act of 2010 (ACA) requires CMS used by MA plans, the net UC payments that hospitals
to redistribute a portion of DSH payments to fund will receive from FFS and from MA plans in 2023 are
hospitals’ uncompensated care costs. Starting in approximately the sum of:
2014, the ACA required CMS to reduce operating DSH
payments to an amount equal to 25 percent of prior- • add-on payments to FFS rates = historical UC costs
law operating DSH payments. The remaining 75 percent × 0.20
is then primarily used to fund a UC pool. (For more
• add-on payments to MA rates = MA discharges
information on Medicare’s DSH and UC costs, see our
× (historical UC costs × 0.20 / historical FFS
June 2022 report to the Congress.) Each DSH hospital
discharges)
receives a share of the FFS UC pool equal to its share
of aggregate UC costs at all DSH hospitals. The ACA Notably, the MA add-on payment formula can be
did not specify the mechanism CMS should use to rewritten as (MA discharge / FFS discharges) ×
distribute these UC funds. (historical UC costs x 0.20). The formulas above
illustrate how hospitals with larger MA shares (i.e., a
96 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
TABLE
3A–1 The MSNI supports hospitals that treat a high share of low-income Medicare beneficiaries,
have high uncompensated care burdens, or serve a high share of Medicare beneficiaries
MSNI
components Weight Data Principle Considerations
LIS share of FFS 1 Average of inpatient • Treating LIS beneficiaries could entail • This component could
Medicare claims and outpatient costs not captured by the MS–DRG or be expanded to include
(inpatient and percentages: APC systems. LIS share of MA claims if/
outpatient) Inpatient • Hospitals that treat a large share of when the encounter data
• FFS inpatient claims LIS beneficiaries tend to have more are sufficiently complete.
for LIS beneficiaries Medicare bad debts (i.e., receive a
(numerator) smaller share of allowed cost sharing).
• FFS inpatient claims Medicare currently pays for 65 percent
for all beneficiaries of these bad debts for FFS Medicare
(denominator) beneficiaries.
Outpatient • Measure is similar to the current
• FFS outpatient DPP measure but includes a broader
claims for LIS measure of low-income Medicare
beneficiaries beneficiaries and does not explicitly
(numerator) include a measure of Medicaid
• FFS outpatient dependence. While Medicaid is not
claims for all explicitly in the formula, hospitals with
beneficiaries high shares of LIS beneficiaries also
(denominator) tend to have higher Medicaid shares.
Uncompensated 1 • Uncompensated • Hospitals that have a high share of • This component implies
care costs as care costs (non- uncompensated costs could face more that Medicare would
a share of a Medicare bad debt financial pressure than the average indirectly subsidize
hospital’s total and charity care) hospital. non-Medicare patients,
revenue (numerator) • Uncompensated care costs could but the effect would be
• Total, all- stem from treating patients without much less direct than
patient revenue insurance or patients with insurance the current system and
(denominator) who cannot afford to pay their would be tied to Medicare
• Data pulled from deductibles or cost sharing. volume.
cost reports
Medicare share 0.5 • MA + FFS hospital • Over the last 25 years, hospitals’ • Outpatient volume could
of all hospital acute inpatient days Medicare margins have shifted be included if/when MA
inpatient days (numerator) from being substantially positive to encounter data are more
• Total, all-payer substantially negative. complete.
hospital inpatient • Therefore, hospitals that • MA days should be
days (denominator) disproportionately treat Medicare included in determining
• Data pulled from beneficiaries face increased risk of safety-net status because
cost reports financial pressure or closure. those hospital patients
• Weight of 0.5 is based on regression have similar costs and
analyses that show the effect of revenue as FFS patients.
Medicare shares on margins is
about half that of the LIS share and
uncompensated care cost measures.
Note: MSNI (Medicare Safety-Net Index), LIS (low-income subsidy), FFS (fee-for-service), MS–DRG (Medicare severity–diagnosis related group), APC
(ambulatory payment classification), MA (Medicare Advantage), DPP (disproportionate patient percentage). The MSNI identifies low-income
beneficiaries as those receiving Part D’s LIS, which includes all beneficiaries who receive full or partial Medicaid benefits, as well as those who do
not qualify for Medicaid benefits in their state but who receive the Part D LIS because they have limited assets and an income below 150 percent of
the federal poverty level. The measure of low-income beneficiaries incorporated in the DPP is limited to inpatient days for Medicare beneficiaries
eligible for Supplemental Security Income; in 2021, these beneficiaries had incomes below 74 percent of the federal poverty level.
Source: MedPAC.
Report to the Congress: Medicare Payment Policy | March 2023 97
TABLE
3A–2
Characteristics of hospitals that benefit from
four different safety-net payment mechanisms
Characteristics of Characteristics of
Safety-net payment policy hospitals that benefit more hospitals that benefit less
Traditional DSH • High Medicare inpatient volume • High level of uncompensated
From 1986 to 2013, there was (the add-on is a Medicare inpatient care relative to Medicare revenue
a percentage add-on to IPPS payments add-on) • Outpatient-focused hospitals
based on each hospital’s (1) Medicaid • High share of Medicaid days
share of total inpatient days and (2) SSI • High share of Medicare patients on
beneficiaries’ share of Medicare inpatient SSI
days.
DSH and temporary UC • High number of Medicaid days • Few Medicaid inpatient days
From 2014 to 2017,* CMS paid hospitals • High number of Medicare SSI days relative to overall Medicare
add-ons to IPPS payments equal to: revenue
• 25% of the traditional DSH payment • Outpatient-focused hospitals
• plus a fixed payment per Medicaid or
Medicare SSI day (e.g., $174/day in 2016).
Both are add-ons to FFS IPPS payments.
Current DSH and UC • High level of UC relative to total • High number of FFS Medicare
• From 2018 to 2020,* CMS transitioned revenue patients but relatively little UC
from the temporary model to the • High levels of MA patients relative
current DSH/UC model. Hospitals to FFS patients
currently receive 25% of traditional DSH.
• Hospitals also currently receive
approximately 20% of uncompensated
costs as an add-on to FFS inpatient
payments.
Illustrative MSNI • High Medicare share of days • High UC burden but few
CMS would pay hospitals add-ons to relative to all inpatient days Medicare patients
Medicare IPPS and OPPS payments based • High share of LIS Medicare claims
on each hospital’s (1) Medicare shares, (2) relative to all Medicare claims
share of their Medicare patients receiving • Outpatient-focused hospitals
LIS benefits, and (3) UC costs relative to benefit more than in the other
total revenue. models
Note: DSH (disproportionate share hospital), IPPS (inpatient prospective payment systems), SSI (Supplemental Security Income), UC (uncompensated
care), FFS (fee-for-service), MA (Medicare Advantage), MSNI (Medicare Safety-Net Index), OPPS (outpatient prospective payment system), LIS (low-
income subsidy). LIS beneficiaries include all beneficiaries who receive full or partial Medicaid benefits, as well as those who do not qualify for
Medicaid benefits in their state but who receive the Part D LIS because they have limited assets and an income below 150 percent of the federal
poverty level.
*The uncompensated care payments in 2019 were still a blend of one-third payment based on Medicaid and SSI days and two-thirds based on
reported uncompensated care costs. Uncompensated care payments were fully based on uncompensated care costs in 2020.
Source: MedPAC.
large MA discharges to FFS discharges ratio) will receive As shown in Table 3A-3, Hospitals A and B both have
a higher share of their UC costs paid by Medicare and historical UC costs of $2 million and will admit 1,000
hospitals with low MA shares will receive a lower share Medicare beneficiaries in 2023. However, Hospital A
of their UC costs paid by Medicare. (For an example, will admit 250 FFS beneficiaries and 750 MA enrollees,
see Table 3A-3.) while Hospital B will admit 750 FFS beneficiaries and
98 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
TABLE
Example of how uncompensated care payments are biased against
3A–3 hospitals in markets with primarily Medicare fee-for-service patients
Illustrative Hospital A Illustrative Hospital B
(higher MA share) (lower MA share)
Historical uncompensated care costs $2 million $2 million
Medicare discharges in 2023
FFS (historical and 2023) 250 750
MA 750 250
Total 1,000 1,000
Uncompensated care payments in 2023
FFS $0.4 million $0.4 million
($2 million × 20%) ($2 million × 20%)
MA (pays FFS prices) $1.2 million $0.13 million
($0.4 million / 250 × 750) ($0.4 million / 750 × 250)
Total FFS + MA $1.6 million $0.53 million
Share of uncompensated care costs paid by Medicare 80% 27%
Note: MA (Medicare Advantage), FFS (fee-for-service). In 2023, disproportionate share hospitals will receive FFS uncompensated care payments equal
to approximately 20 percent of their historic uncompensated care costs. Based on the literature and staff discussions with insurers and hospital
systems, we assume that MA plans pay hospitals rates approximately equal to FFS rates.
Source: MedPAC.
250 MA enrollees. However—despite having equal UC To avoid the payment rate distortions and MA
costs—the second hospital receives about one-third benchmark distortions caused by the current system
the UC payments received by the first hospital because for distributing UC payments, the Commission has
the second hospital’s FFS share of discharges is one- repeatedly suggested that CMS pay hospitals directly
third the first hospital’s FFS share. The system is set up for a portion of their UC costs and that the MA plans’
to make sure that FFS Medicare pays each DSH hospital portion also be paid directly by the Medicare program
an equal share of its UC costs, but current policy’s (Medicare Payment Advisory Commission 2022b,
unintended consequence is that MA plans can pay each Medicare Payment Advisory Commission 2013). ■
hospital vastly different shares of their UC costs. Larger
ratios of MA discharges to FFS discharges result in
higher total UC payments.
Report to the Congress: Medicare Payment Policy | March 2023 99
Endnotes
1 Throughout this chapter, we use the term “FFS Medicare” as 8 Hospital closures are defined as cessation of Medicare
equivalent to the CMS term “Original Medicare.” beneficiaries’ access to inpatient services at a general
short-term ACH or CAH in the United States (exclusive
2 For example, Medicare pays separately for general acute of territories). Closures do not include the relocation of
care hospitals’ facility costs for services provided in hospital- inpatient services from one hospital to another under
based psychiatric units, post-acute care units, and clinics. common ownership within 10 miles, nor do closures include
Medicare also pays separately (outside of the IPPS and OPPS) hospitals that both opened and closed within a 5-year time
for hospitals’ direct costs of graduate medical education, as period. The number of hospital closures and openings in a
well as organ acquisition. These other Medicare payment given year can change over time as hospitals reopen or dates
methodologies are beyond the scope of this chapter. of closure are updated.
3 Unless otherwise noted, all years referring to inpatient 9 We measure closures during each fiscal year (to match the
services refer to fiscal year while those referring to timeframe of Medicare inpatient payment rate changes).
outpatient services refer to calendar year, consistent with Other sources such as the University of North Carolina
when CMS updates these payment systems. For more details measure closures during each calendar year.
on the IPPS and OPPS, see Hospital Acute Inpatient Services
Payment System at https://www.medpac.gov/wp-content/ 10 Small rural hospitals are eligible to convert to REHs, which
uploads/2021/11/MedPAC_Payment_Basics_22_hospital_ provide 24/7 emergency services and other outpatient
FINAL_SEC.pdf and Outpatient Hospital Services Payment services. REHs receive a monthly fixed rate for their standby
System in our Payment Basics series at https://www.medpac. costs, enhanced outpatient rates, and standard rates for
gov/wp-content/uploads/2021/11/MedPAC_Payment_ other services.
Basics_22_OPD_FINAL_SEC.pdf.
11 Since mid-2020, hospitals have had to report weekly
4 Under the IPPS and OPPS, FFS Medicare pays the prospective occupancy data to the Department of Health and Human
rate minus any beneficiary cost-sharing responsibilities Services. According to these data, general ACHs’ occupancy
(which the provider collects from the beneficiary or a rates in fiscal years 2021 and 2022 were about 74 percent and
supplemental insurer). Medicare reimburses hospitals 77 percent, respectively, with some hospitals exceeding 90
for 65 percent of bad debts resulting from beneficiaries’ percent in certain months. These higher occupancy rates
nonpayment of cost sharing after hospitals have made may more accurately reflect decreases in staffed beds during
reasonable efforts to collect the unpaid amounts. part of the year, but they may also reflect inaccurate data
from some hospitals.
5 For more details on the CAH payment system, see
Critical Access Hospitals Payment System in our Payment 12 The decline in the number of inpatient stays (–6.1 percent)
Basics series at https://www.medpac.gov/wp-content/ was larger than the decline in stays per beneficiary (–1.8
uploads/2021/11/MedPAC_Payment_Basics_22_CAH_ percent) because there was a 4.3 percent decline in FFS
FINAL_SEC.pdf. beneficiaries (as a greater share of Medicare beneficiaries
enrolled in MA).
6 Medicare uses the OPPS to pay for the facility costs of
outpatient services at post-acute care hospitals (i.e., long- 13 CAHPS® is a registered trademark of the Agency for
term care and rehabilitation hospitals), at certain specialized Healthcare Research and Quality.
short-term acute care hospitals (i.e., psychiatric, cancer,
and children’s hospitals), and at community mental health 14 The all-payer operating margin at CAHs was also a record
centers. high of 10.8 percent (6.0 percent exclusive of federal relief
funds).
7 While the focus of this chapter is on assessing the adequacy
of IPPS and OPPS payments, we include all general ACHs 15 The federal relief funds are primarily from the Provider
(defined as those paid under the IPPS as well as CAHs and Relief Fund but in 2020 also included forgiven loans from the
ACHs in Maryland and in U.S. territories) in our indicators of Paycheck Protection Program.
beneficiaries’ access to care because all general ACHs provide a
range of acute hospital inpatient and outpatient services to FFS 16 The increase in IPPS hospitals’ operating margin in 2021
Medicare beneficiaries and therefore can serve as substitutes was even larger prior to the inclusion of relief funds (over
for care at general ACHs paid under the IPPS and OPPS. 5 percentage points), as hospitals’ revenue prior to relief
100 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
funds increased faster (over 13 percent). The rapid revenue effect). In 2021, estimated DSH payments decreased about
growth from 2020 reflects the low 2020 level of operating 9 percent while uninsured rates increased by slightly less.
revenue (excluding relief funds). IPPS hospitals’ all-payer However, as sequestration was suspended for all of 2021 but
total margin had an even larger increase in 2021, growing only part of 2020, the net effect was a minimal change in
over 4 percentage points (near 6 percentage points excluding Medicare’s uncompensated care payments to IPPS hospitals.
relief funds), reflecting strong investment returns.
23 The OPPS also applies budget-neutrality factors to the base
17 We calculated aggregate operating margins net of interest rate; however, these offset the estimated effects of other
expense, before taxes and extinguishment of debt, based on policy changes (such as updated geographic adjustments and
hospital systems’ financial reports. pass-through payments) and therefore should not affect total
payments.
18 “Same-hospital net patient service revenue per adjusted
admission decreased 4.2 percent year-over-year for third 24 We estimated that the effect of the suspension of the
quarter 2022, primarily due to lower COVID-19-related acuity 2 percent sequestration on Medicare program OPPS
and lower COVID-19 volumes, partially offset by improved payments per beneficiary in calendar year 2021 was a 0.6
pricing yield. COVID-19 admissions were 6 percent of total percent increase relative to 2020 for two reasons. First,
admissions in the third quarter of 2022 versus 10 percent in the suspension was in effect for all of calendar year 2021,
the third quarter of 2021” (Tenet Health 2022). compared with eight months of calendar year 2020. Second,
the 2 percent sequestration does not apply to FFS Medicare
19 Given that hospitals with distinct units can affect the margin beneficiary cost sharing, which is about 20 percent of all
of inpatient and outpatient service lines based on where OPPS payments.
they treat patients (e.g., having a SNF in the hospital may
allow earlier discharges from the inpatient unit), we focus 25 If costs would have been reduced more in the absence of
our Medicare margin discussion on hospitals’ aggregate relief funds, the margin decline would have been smaller.
Medicare margin across multiple hospital service lines
(including inpatient, outpatient, swing bed, skilled nursing, 26 We have also found that hospitals under financial pressure
rehabilitation, psychiatric, and home health services) as well (those that do not have material profits on non-Medicare
as direct graduate medical education and uncompensated patients) have a stronger incentive to control costs and
care payments. roughly broke even in Medicare in recent years. For-profit
hospitals, which have an incentive to maximize shareholder
20 Because federal relief funds were intended to help cover returns, have also roughly broken even in Medicare in recent
lost revenue and payroll costs—including lost revenue from years (Medicare Payment Advisory Commission 2021).
Medicare patients and the cost of staff who help treat these
patients—we include a portion of these relief funds (based 27 The objective of this analysis is to find a subset of the
on FFS Medicare’s share of 2019 all-payer operating revenue) relatively efficient hospitals rather than to identify all
in our Medicare margins. Using this method, we allocated efficient hospitals. For example, we exclude from our
$3.5 billion of the $18 billion in federal funds that hospitals analysis small hospitals with less than 500 inpatient stays,
reported on their cost reports with midpoints in fiscal year not because we know they are inefficient but because we
2021 toward hospitals’ care of FFS Medicare beneficiaries. have an insufficient volume of claims to know whether they
performed at a relatively efficient level.
21 We estimated that the effect of the suspension of the
2 percent sequestration on IPPS payments per stay in fiscal 28 We use medians rather than means to limit the influence of
year 2021 was a 1.1 percent increase relative to 2020 for two outliers on our set of efficient providers.
reasons. First, the suspension was in effect for all of fiscal
year 2021, compared with five months of fiscal year 2020. 29 We do not adjust our costs per inpatient stay for economies
Second, the 2 percent sequestration does not apply to FFS of scale. However, we excluded all hospitals with fewer than
Medicare beneficiary cost sharing, which is about 20 percent 500 Medicare inpatient stays from our analysis. For the
of all IPPS payments. remaining hospitals, economies of scale are not a material
factor when evaluating costs per discharge because costs are
22 Under current law, aggregate uncompensated care payments roughly proportionate to the volume of stays for hospitals
are set prospectively by CMS as the product of two estimates with over 500 Medicare stays per year (generally over 1,000
for the upcoming payment year: 75 percent of DSH payments all-payer stays). Teaching hospitals tend to have higher costs
under prior law and the uninsured rate as a percentage of the per stay, but we standardize costs per stay by adjusting for
rate in 2013. Like other Medicare payments, uncompensated the effect of case mix, outlier cases, and the cost of training
care payments are subject to sequestration (when it is in residents. After these adjustments, teaching hospital costs,
Report to the Congress: Medicare Payment Policy | March 2023 101
on average, are similar to nonteaching hospital costs. For a 34 The DSH metric we use is called the DSH patient percentage,
more complete description of the methodology, see online which is the sum of two ratios: Medicare Supplemental
Appendix 3-B from our 2016 report to the Congress, available Security Income (SSI) days (for MA and FFS patients) as a
at http://www.medpac.gov. share of all Medicare days, and days in which Medicaid is a
primary payer as a share of all inpatient days.
30 We adjust costs per stay for the share of Medicare patients
that are on Supplemental Security Income (SSI). However, 35 Starting at the 10th percentile would result in about 170
we do not adjust readmission or mortality metrics for patient hospitals that currently receive DSH and uncompensated
income, in keeping with our policy of not adjusting quality care payments not receiving any Medicare safety-net
metrics for income. The efficient group of hospitals tends to payments. Alternatively, the policy could start at the 5th
have a smaller share of low-income Medicare beneficiaries. percentile to reduce the number of hospitals that currently
The median share of Medicare patients on SSI for efficient receive DSH payments but would then receive no MSNI
hospitals is 6.3 percent and the median share for other payments.
hospitals is 7.8 percent. In 2022, we also developed a measure
of Medicare safety-net status called the Medicare Safety-Net 36 While the total amount of DSH and uncompensated care
Index (MSNI) (see the appendix to this chapter). Hospitals payments distributed to hospitals in 2019 (after accounting
in the highest quartile of the MSNI metric (the sum of the for sequestration) was about $11.7 billion, the amount
share of Medicare beneficiaries receiving Part D’s low-income distributed to the hospitals in our simulation was $11.2 billion.
subsidy (LIS) benefit, one-half the Medicare patient share, Not all hospitals were included in our simulation because
plus uncompensated care costs divided by total revenues) some hospitals had incomplete data to create the MSNI or did
are half as likely to be in our efficient-provider group. These not meet our simulation requirement of having a minimum of
hospitals tend to have low costs, but they perform worse on 500 FFS discharges.
mortality and readmission metrics, possibly due to their small
size and the fact that high-MSNI hospitals have higher shares 37 Under this model of the MSNI, certain hospitals that
of low-income Medicare beneficiaries. currently are paid based on historical costs as a sole
community hospital (SCH) or partially based on historical
31 CMS is still evaluating how to apply the Supreme Court’s costs as a Medicare-dependent hospital (MDH) would not
decision in American Hospital Association v. Becerra, 142 S. see an FFS shift in revenue. Under an MSNI policy, these
Ct. 1896 (2022) to 2018–2012. In the 2023 OPPS final rule, CMS hospitals would receive the higher of the current MDH or
addressed the court ruling only for 2023 by reverting the SCH payments based on historical costs or the benefit of the
payments for 340B drugs back to average sales price plus 6 MSNI adjustment. Therefore, the benefits to rural hospitals
percent (the rate prior to 2018, when CMS lowered the rate to are conservatively stated in this chapter.
average sales price minus 22.5 percent).
38 Currently, government-owned hospitals’ DSH and
32 Our low-income definition (those receiving the LIS in Part uncompensated care payments in some cases can result in
D) is much more inclusive than the current definition of low- an 80 percent or larger increase in their Medicare payments
income patients used in the current disproportionate share relative to standard Medicare rates.
formula (those receiving Supplemental Security Income (SSI)).
Our prior work indicated that the LIS variable was a better 39 We tested to see if the difference was due to rural areas, but
predictor of closure and margins than the SSI variable, but even limiting the analysis to urban hospitals, hospitals in
they are positively correlated (correlation coefficient of 0.81). markets with lower MA penetration tended to benefit more
from the transition to the MSNI.
33 The add-on amount determines the amount in the FFS
Pricer and thus most MA payment amounts. The add-on 40 In 2021, about 46 percent of beneficiaries with Part A and
is seen as an interim payment amount. At the end of the Part B were enrolled in MA plans, up from 26 percent in 2010
year, CMS makes additional adjustments to uncompensated (Medicare Payment Advisory Commission 2021).
care payments so that each hospital receives a share of
uncompensated care payments equal to that hospital’s share 41 Hospital representatives have said that MA plans typically pay
of all DSH hospitals’ uncompensated care costs. Thus, the FFS rates for hospital care but that hospitals (1) receive lower
FFS Pricer amount (the Pricer is software CMS uses to set total payments from MA plans because of increased rates of
interim FFS Medicare prices and that MA plans in turn base medical necessity denials by MA plans and (2) incur higher
their prices on) does not determine the final amount of administrative costs for MA beneficiaries relative to FFS
uncompensated care payments received by DSH hospitals. In beneficiaries because of prior authorization processes put in
contrast, MA payments are determined by the FFS Pricer and place by MA plans. We cannot quantify these effects.
are not reconciled at the end of the year.
102 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
42 The claims could be priced using FFS prices. amount to reflect declines in the share of the population
that is uninsured). This computation yields a fixed pool of
43 The Medicaid program does not have a definition of “safety- FFS UC dollars to be distributed among DSH hospitals. Each
net” hospital. However, hospitals with a Medicaid inpatient DSH hospital’s share of that pool of dollars is equal to that
utilization rate at least one standard deviation above the hospital’s historical UC costs divided by the aggregate of all
mean for the state are “deemed DSH hospitals” and states DSH hospitals’ historical UC costs.
are required to provide some Medicaid DSH payments to
those hospitals (Medicaid and CHIP Payment and Access The uncompensated care payments due to each hospital
Commission 2022a). are then divided by that hospital’s historical number of FFS
discharges to arrive at a per discharge add-on amount. That
44 The law exempts the 3 percent of hospitals with the highest hospital-specific add-on amount is added to each hospital’s
number and share of patients who are eligible for Medicare inpatient payment rate. To the extent that actual FFS
and receive Supplemental Security Income (SSI) from this discharges differ from this historical number, the difference
change (Medicaid and CHIP Payment and Access Commission is reconciled in the final cost report settlement each year, to
2022a). ensure that each hospital receives exactly the amount it is
due as published in that year’s final IPPS rule.
45 There is also a secondary effect. As FFS inpatient stays
increase, UC payments per stay decrease, which creates a 47 For 2023, the historical UC costs are based on the average
decrease in MA prices paid per unit of service. Shifting to the of 2018 and 2019 cost reports, and historical FFS discharges
MSNI would remove this distortion. are the average of 2018, 2019, and 2021 FFS discharges. In
addition, starting in 2023, DSH hospitals in Puerto Rico and
46 In each year’s IPPS final rule, CMS first computes the FFS hospitals administered by the Indian Health Service will
UC pool as an amount equal to 75 percent of what DSH receive supplemental UC payments.
payments would have been under prior law (reduced by an
Report to the Congress: Medicare Payment Policy | March 2023 103
References
Adegbesan, A. 2022. Hospitals to lean on more expensive Community Health Systems. 2020. Quarterly report (Form 10-Q).
travel nurses even after Covid. Bloomberg News, March 15. Filing submitted to the Securities and Exchange Commission.
https://www.bloomberg.com/news/articles/2022-03-15/ https://chsnet.gcs-web.com/node/20561/html.
budget-busting-travel-nurses-to-rack-hospitals-even-after-
covid?leadSource=uverify%20wall. Government Accountability Office. 2013. VA health care:
Management and oversight of fee basis care need improvement.
Ascension. 2022a. Consolidated financial statements and GAO-13-441. Washington, DC: GAO.
supplementary information years ended June 30, 2022 and
2021, with reports of independent auditors. https://ascension. HCA Healthcare. 2022. Quarterly report (Form 10-Q). Filing
org/-/media/Files/Ascension/About/Community-Investor- submitted to the Securities and Exchange Commission. https://
Relations/2022/Consolidated-Ascension-Financial-Statements- s23.q4cdn.com/949900249/files/doc_financials/2022/
Q4-FY22.pdf?la=en&hash=17A905C173689E9F8323F399A83A85B7 q3/39af3594-88b0-4b14-934f-bf578787127f.pdf.
6DF61117.
HCA Healthcare. 2020. Quarterly report (Form 10-Q). Filing
Ascension. 2022b. Management’s discussion and analysis of submitted to the Securities and Exchange Commission. https://
financial condition and results of operations for Ascension, sec.report/Document/0001193125-20-281070/.
as of and for the years ended June 30, 2022 and 2021. https://
ascension.org/-/media/Files/Ascension/About/Community- Maeda, J., and L. Nelson. 2017. An analysis of private-sector prices
Investor-Relations/2022/Ascension-Management-Discussion- for hospital admissions. Congressional Budget Office working
and-Analysis-Q4-FY22.pdf?la=en&hash=81465554EBD0A72F3FF0 paper 2017-02. Washington, DC: CBO.
DB91EC6B8954CD8A61FC.
Medicaid and CHIP Payment and Access Commission. 2022a.
Ascension. 2020. Consolidated financial statements and Report to the Congress on Medicaid and CHIP. March. Washington,
supplementary information years ended June 30, 2020 and DC: MACPAC.
2019, with reports of independent auditors. https://ascension.
org/-/media/Files/Ascension/About/Community-Investor- Medicaid and CHIP Payment and Access Commission. 2022b.
Relations/2020/Consolidated-Ascension-Financial-Statements- Report to the Congress on Medicaid and CHIP. June. Washington,
Q4-FY20-updt.pdf?la=en&hash=83394A82F76D24CF069CF2ECD2 DC: MACPAC.
388F24AE8747B2.
Medicare Payment Advisory Commission. 2022a. A data book:
Berenson, R. A., J. H. Sunshine, D. Helms, et al. 2015. Why Health care spending and the Medicare program. Washington,
Medicare Advantage plans pay hospitals traditional Medicare DC: MedPAC. https://www.medpac.gov/wp-content/
prices. Health Affairs 34, no. 8 (August): 1289–1295. uploads/2022/07/July2022_MedPAC_DataBook_SEC_v2.pdf.
Centers for Medicare & Medicaid Services, Department of Health Medicare Payment Advisory Commission. 2022b. MedPAC
and Human Services. 2013. Medicare program; hospital inpatient comment on CMS’s proposed rule on the hospital inpatient
prospective payment systems for acute care hospitals and the prospective payment system and the long-term care hospital
long term care; hospital prospective payment system and fiscal prospective payment system for FY 2023. June 16. https://
year 2014 rates; quality reporting requirements for specific www.medpac.gov/wp-content/uploads/2022/06/06162022_
providers; hospital conditions of participation; payment policies FY2023_IPPS_LTCH_MedPAC_COMMENT_v2_SEC.pdf.
related to patient status. Final rule. Federal Register 78, no. 160
(August 19): 50495–51040. Medicare Payment Advisory Commission. 2022c. Report to
the Congress: Medicare and the health care delivery system.
CommonSpirit. 2022. Unaudited annual report for the years Washington, DC: MedPAC.
ended June 30, 2022 and 2021. https://www.commonspirit.org/
content/dam/commonspirit/pdfs/2022-CommonSpirit-Health- Medicare Payment Advisory Commission. 2022d. Report to the
MDA-and-Financial-Statements-SECURED.pdf. Congress: Medicare payment policy. Washington, DC: MedPAC.
Community Health Systems. 2022. Community Health Systems, Medicare Payment Advisory Commission. 2021. Report to the
Inc. announces third quarter ended September 30, 2022 results. Congress: Medicare payment policy. Washington, DC: MedPAC.
October 26. https://chsnet.gcs-web.com/news-releases/news-
release-details/community-health-systems-inc-announces- Medicare Payment Advisory Commission. 2019. Report to the
third-quarter-ended. Congress: Medicare payment policy. Washington, DC: MedPAC.
104 H o s p i t a l i n p a t i e n t a n d o u t p a t i e n t s e r v i c e s : A s s e s s i n g p a y m e n t a d e q u a c y a n d u p d a t i n g p a y m e n t s
Medicare Payment Advisory Commission. 2013. Comment Scott, D. 2021. The public option is now a reality in 3 states. Vox,
letter to CMS on the proposed rule entitled: Medicare program; June 17. https://www.vox.com/policy-and-politics/22535267/
hospital inpatient prospective payment systems for acute care public-option-health-insurance-nevada-colorado-washington.
hospitals and the long-term care hospital prospective payment
system and proposed fiscal year 2014 rates; quality reporting Tenet Health. 2022. Tenet reports third quarter 2022 results;
requirements for specific providers; hospital conditions of announces $1 billion share repurchase program. https://s23.
participation, June 25. https://www.medpac.gov/wp-content/ q4cdn.com/674051945/files/doc_financials/2022/q3/THC-
uploads/import_data/scrape_files/docs/default-source/ 2022.09.30-EarningsRelease.pdf.
comment-letters/medpac-s-comment-on-cms-s-acute-and-
long-term-care-hospitals-proposed-rule.pdf. Tenet Health. 2020. Quarterly report (Form 10-Q). Filing
submitted to the Securities and Exchange Commission. https://
Medicare Payment Advisory Commission. 2011. Report to the sec.report/Document/0000070318-20-000043/.
Congress: Medicare payment policy. Washington, DC: MedPAC.
Trinity Health. 2022a. Consolidated financial statements as of
Moody’s Investors Service. 2022. 2023 Outlook: Negative and for the years ended June 30, 2022, and 2021, supplemental
as inflation, labor costs continue to drive expenses higher. consolidating schedules as of and for the year ended June 30,
December 7. https://www.moodys.com/research/Not-For- 2022, and independent auditor’s reports. https://www.trinity-
Profit-and-Public-Healthcare-US-2023-Outlook-Negative-- health.org/assets/documents/financials/trinity-health-fy22-
PBC_1342403. financial-statements-long-form-final.pdf.
Nguyen, N. X., and S. H. Sheingold. 2011. Indirect medical Trinity Health. 2022b. Trinity Health’s proactive response
education and disproportionate share adjustments to Medicare to healthcare pandemic aftershocks significantly mitigated
inpatient payment rates. Medicare & Medicaid Research Review 1, operating losses and strengthens the system for strategic growth
no. 4: E1-E19. for the year ended June 30, 2022. https://www.trinity-health.
org/assets/documents/financials/trinity-health-mda-ye-jun-
S&P Global. 2022. S&P municipal bond hospital index. November fy22-final-posted-2022.09.30.pdf.
10. https://www.spglobal.com/spdji/en/indices/fixed-income/
sp-municipal-bond-hospital-index/#overview. Trinity Health. 2020. Consolidated financial statements as of
and for the years ended June 30, 2020, and 2019, supplemental
S&P Global Ratings. 2022. Outlook for U.S. not-for-profit acute consolidating schedules as of and for the year ended June 30,
health care: A long road ahead. December 1. https://www. 2020, and independent auditor’s reports. https://www.trinity-
spglobal.com/ratings/en/research/articles/221201-outlook- health.org/assets/documents/financials/trinity-health-fy20-
for-u-s-not-for-profit-acute-health-care-a-long-road- financial-statement-with-exhibits.pdf.
ahead-12573554.
White, C., A. M. Bond, and J. D. Reschovsky. 2013. High and
Schramm, S., and Z. Aters. 2021. Estimating the impact of varying prices for privately insured patients underscore hospital
reference-based hospital pricing in the Montana state employee market power. HSC research brief no. 27. Washington, DC: Center
plan. National Academy for State Health Policy. https://www. for Studying Health System Change.
nashp.org/wp-content/uploads/2021/04/MT-Eval-Analysis-
Final-4-2-2021.pdf.
Report to the Congress: Medicare Payment Policy | March 2023 105
File and source
- File
- hospital-inpatient-and-outpatient-services-march-2023-report-to-the-congress-ch-.pdf
- Size
- 971,997 bytes
- SHA-256
- 83db294897526fce429cbf2c647f3218afbb4c991746949e09749202809b21d0
- Original
- www.medpac.gov