DEPARTMENTAL GRANT APPEALS BOARD
Department of Health and Human Services
SUBJECT: Michigan Department of Social Services
Docket No. 86-209
Audit Control No. 05-65460
Decision No. 872
DATE: June 1, 1987
DECISION
The Michigan Department of Social Services (State) appealed a decision
by
the Health Care Financing Administration (Agency) disallowing
$264,318 in
federal financial participation (FFP) claimed by the State
under Title XIX of
the Social Security Act (Act) during the period
October 1, 1977, through
September 30, 1983. The State later accepted
the Agency's determination
with respect to $57,242 of the disallowance
(appeal brief at 1-2), so that
$207,076 remains in dispute. This amount
represents the State's claims
for the Title XIX (Medicaid) share of
costs incurred by the State survey
agency in inspecting general and
psychiatric hospitals for compliance with
State, Medicaid, and Medicare
(Title XVIII of the Act) requirements.
The State maintained that the method adopted by the Agency in
allocating
costs between Medicaid and Medicare for the facilities involved
was
faulty because it was internally inconsistent. The Agency allocated
all
of the federal share of the costs of surveying accredited general
and
psychiatric hospitals that participated in both Medicaid and
Medicare
solely to Medicare, while the federal share for nonaccredited
facilities
was split equally between the two programs.
Upon review of the record in this proceeding, we conclude that the
Agency
has offered no rationale in support of the reasonableness of its
allocation
methodology in response to the State's challenge. The
methodology is
internally inconsistent, and the Agency has offered no
reasonable explanation
for this inconsistency. Moreover, the Agency's
approach is contrary to
Agency policy. Consequently, the disallowance
cannot stand.
Background
Under section 1864 of the Act, 42 U.S.C. 1395aa(a), the Agency enters
into
agreements with states whereby the state survey agency is used to
survey
certain entities, including general and psychiatric hospitals, to
determine
whether they meet the statutory and regulatory requirements
for participation
in Medicare. See 45 CFR 405.1901 et seq. The state
survey agency
is reimbursed by Medicare for these functions. 42
U.S.C.
1395aa(b). The Act also provides that the same state survey
agency will
perform Medicaid standards surveys for the State Medicaid
agency. 42
U.S.C. 1396b(a)(4). Medicaid regulations provide that
FFP is available
for expenditures that the state survey agency makes to carry
out these
Medicaid survey and certification functions. 42 CFR
431.610(h)(1). FFP
is not available for agency expenditures that are
attributable to the
survey agency's overall responsibilities under state law
and regulations
for establishing and maintaining standards. 42 CFR
431.610(h)(2).
In the background of this dispute is an audit by a private auditing
firm
on contract with the Office of Inspector General (OIG) in which
the
auditors examined the allocation of survey agency payroll hours
to
Medicare, Medicaid, and the State. The state's system of
calculation,
which it called a participative responsibility distribution
(PRD)
system, involved a three step process. First, payroll hours
were
grouped by facility type (i.e., hospital, nursing home, etc.).
Second,
the facilities were sorted by licensure characteristic (i.e.,
Medicaid,
Medicare and State survey, Medicare and State survey, Medicaid and
State
survey, and State only survey), and survey agency payroll hours
which
were kept by facility type were allocated based on
licensure
characteristic. Third, the calculated percentages were
distributed to
the various programs based on the estimated time requirements
of each
facility and licensure type. It is undisputed that the
licensure
characteristic percentages used for the facilities in dispute
here
(general and psychiatric hospitals surveyed for all three programs)
were
based upon a 1972 agreement between the State and the Agency.
The auditors found that the State had made several errors in
calculating
its claims for FFP. Among the most serious of these errors
was that
facilities which were not participating in the Medicare and
Medicaid
programs (nonparticipating facilities) but were surveyed for
State
licensure were not included in the listing of facility type by
licensure
characteristic, so that the allocations to Medicare and Medicaid
were
overstated. The auditors recalculated the payroll hours and
adjusted
the percentages to reflect the inclusion of previously
overlooked
categories of facilities. This recalculation resulted in a
substantial
shifting of expenses claimed by the State under Medicaid to the
Medicare
program.
The State agreed with the auditors' findings concerning the errors in
its
allocation system. 1/ In fact, it had come to the same
conclusions
in a self-review performed in 1983 and had submitted a
revised
participative responsibility distribution (PRD) system in its
fiscal
year (FY) 1984 budget request that was intended to rectify
these
problems. The Agency accepted the revised PRD for FY 1984 and
for
subsequent years through 1987.
The dispute here arises because, although the State substantially
agreed
with the adjustments made by the auditors to the number of
survey
payroll hours, the State did not agree with the percentages used by
the
auditors in calculating the appropriate charges to the two
federal
programs. The auditors used the State's revised PRD for
categories of
hospitals which had not previously been specifically listed in
the 1972
agreement, and therefore split the federal share of survey costs
evenly
between Medicare and Medicaid. For the two categories included
in the
original 1972 PRD agreement, general and psychiatric hospitals
which
were surveyed for both federal and State purposes, the auditors used
the
methodology of the 1972 agreement, which allocated all of the
federal
share to Medicare. In other words, the auditors split federal
costs
evenly between Medicare and Medicaid for nonaccredited
and
nonparticipating general and psychiatric hospitals, but allocated all
of
the federal share to Medicare for accredited general and
psychiatric
hospitals. As a result, the Agency concluded that while the
State had
apparently underclaimed FFP for the Medicare program by $3,293, it
had
overclaimed in the Medicaid program by $207,076. Due to the fact
that
the State had already expended all of the funds it was due to
receive
under Medicare, however, the shifting of costs from Medicare to
Medicaid
resulted in the State having to pay the increased Medicare
allotment.
The State argued that the auditors' methodology distorted the
allocation
of payroll hours through inconsistent adjustments. It
maintained that
the Agency should either have used only the PRD percentages
set forth in
the original 1972 PRD agreement or all of the revised PRD
percentages
accepted by the Agency in 1984. The State contended that the
auditors'
hybrid methodology is inconsistent and inaccurate. The
State argued
that if it was appropriate to split the federal share of costs
for
nonaccredited general and psychiatric hospitals between Medicare
and
Medicaid, as the auditors did, then it was also appropriate to split
the
costs for accredited general and psychiatric hospitals. The
State
proposed that the Board hold that either the original or the
revised
PRD, but not the auditors' combined "hybrid" methodology, must
be
applied to the revised payroll hours and facility count for
the
disallowance period to calculate the appropriate level of FFP in
this
case. The use of either method would nullify the disallowance.
The Agency's explanation was that the auditors were not inconsistent
in
using percentages from both the 1972 PRD agreement and the
State's
revised PRD percentages from the 1984 budget. The Agency
maintained
that rather than adopting the percentages from the revised PRD
for
facilities that were not covered by the original PRD, the auditors
could
have rejected the entire amount based on lack of documentation.
The
Agency also argued that the State should be bound by the original
PRD
allocations for accredited hospitals, which were based on an analysis
of
time studies undertaken by the State. Agency Ex. 1, p. 1. The
Agency
also contended that the Board had held that the Agency was not
obliged
to apply a revised PRD to previous years, citing Wisconsin Department
of
Health and Social Services, Decision No. 534, April 29, 1984.
Analysis
The issues here are (1) whether the Agency was bound by the 1972
PRD
agreement to accept the State's original accounting for the survey
costs
and, (2) whether the Agency can show a reasonable basis for
the
apparently inconsistent approach which it adopted. We conclude that
the
Agency is not bound by the 1972 PRD agreement; however, the Agency
has
not made the required showing in support of its methodology.
We do not agree with the State that the PRD agreement binds the Agency
to
use PRD percentages calculated using only the categories of
facilities used
by the State over the years. The State provided no
evidence of a
written agreement, or even oral understanding, which would
require this
result. A 1975 description of the State's process,
submitted as Agency
Exhibit 1, refers to agreements on program
percentages, giving accredited
hospitals as an example (State share 90%,
Medicare share 10%) and stating
that a "similar approach was used in all
other types of facilities in the
licensure/certification program."
Agency Ex. 1, p. 2. This
approach is illustrated in a list of
facilities but nothing in the
description of the process indicates that
the parties had agreed to use only
the categories listed; indeed, the
implication is that any other type of
facility joining the
licensure/certification program would bear its share of
survey costs.
In fact, the State itself modified the list to account for
validation
surveys, a workload requirement benefiting only Medicare.
The State
admitted that it had nonetheless calculated its PRD percentages for
the
period in question here using a list which omitted certain types
of
facilities surveyed by the State during that period. In our view,
the
Agency properly modified the list to include all relevant categories
of
facilities in its calculations.
With respect to the State's alternative contention that the Board
should
rule that the revised PRD should be applied to all facilities, we
agree
with the Agency that mere approval of the revised PRD for FY 1984
does
not necessarily mean that that PRD had to be adopted for previous
years.
The issue here, however, is can the Agency, having adopted the
revised
system for some categories, explain why that system should not
be
applied consistently for all categories? The State contended that
the
Agency should use the entire revised PRD because it is more accurate
and
is consistent in its allocation of the federal share between
Medicare
and Medicaid for all categories of facilities.
The Agency did not explain why its inconsistent treatment of the
costs
here -- sometimes splitting the federal share between Medicare
and
Medicaid and sometimes allocating it entirely to Medicare -- is
either
accurate or appropriate. Moreover, the Agency did not cite, and we
could
not find, any basis in the statute or regulations for the
distinction
drawn between accredited and nonaccredited hospitals with respect
to the
relative benefits of survey costs for Medicaid and Medicare.
In the background section of the report, the auditors stated that
"[a]
memorandum from the H.H.S. Office of Inspector General, dated January
2,
1981, recommended that in the absence of acceptable evidence to
the
contrary, the State's share of joint costs should be 33 1/3 percent
for
facilities that participate in both Medicare and Medicaid and 50%
for
facilities that participate in only one of these programs."
Audit
report at 12. This clearly implies that costs should be split
between
the federal programs where both participate. This was not the
method
adopted by the auditors, however.
The auditors adopted the 1972 agreement's method of allocating all of
the
federal share of survey costs for accredited hospitals to Medicare
because
they thought they were bound by the agreement, not because they
agreed with
it. 2/ The auditors also apparently felt bound to adopt the
revised PRD
for other facilities, since it had been accepted by the
Agency for years
subsequent to the disallowance period. The auditors
then refused to
change the original allocation percentages for hospitals
to make them
consistent with the revised method's split between Medicare
and Medicaid,
because, they stated, the State's "proposed retroactive
adjustment of
apparent inappropriate percentage distributions does not
address other
possible inappropriate distributions." Audit report at
12. This
was an apparent reference to the revised PRD's allocation of
90 percent to
federal programs for long-term care facilities. This
seems more an
expression of vexation, however, than an explanation for
the adoption of a
particular methodology.
Agency counsel similarly did not offer an explanation for either
the
internal inconsistency of the PRD adopted as the basis for
the
disallowance or the adoption of a portion of the original PRD by
the
auditors, even though all involved agreed that the original PRD
was
flawed. The Agency tried to justify its use of the original
allocation
for accredited hospitals by stating that the original PRD was
based on
time studies. But the Agency's own exhibit shows that the
rationale for
allocating all of the federal share to Medicare was that
"Medicaid
follows Medicare automatically." Agency Ex. 1, p. 2.
This was true for
the period under the revised PRD as well, however, and also
true for
skilled nursing facilities, where the costs were evenly split
between
the two federal programs under both the original and the revised
PRD
systems.
Rather than address the inconsistency of the methodology, the
Agency
argued that the manner in which it was formulated was not
inconsistent:
where there were allocation percentages for a category of
facility in
the original PRD system, the auditors used them; where there were
no
allocation percentages available because the original PRD system did
not
have them, the auditors used the allocations from the revised
PRD
system. This does not satisfactorily explain why, if it was correct
to
split the federal share for some categories of facilities, this
split
should not be adopted consistently.
The Agency also did not address the State's assertions about
the
shortfalls in the original PRD. Most importantly, the Agency
left
unchallenged the State's claim that the revised PRD was the
most
accurate of the three methodologies, probably because the Agency
had
adopted that very methodology for later periods. While we agree,
as
noted above, that the Agency was not obliged, due merely to
its
acceptance of the revised PRD for subsequent years, to adopt
that
methodology for the disallowance period, the Agency has not provided
a
logical rationale for using the hybrid, inconsistent methodology that
it
did. Given the Agency's apparent policy favoring an equal split
between
programs, which was noted by the auditors, this unsupported
methodology
cannot be sustained.
Accordingly, the disallowance is overturned.
______________________ Donald F. Garrett
_______________________ Norval D.
(John)
Settle
_____________________ Judith A.
Ballard
Presiding Board Member
1. The State explained that it disagreed with a few
minor facility
counts determined by the auditors, but did not dispute those
counts for
the purposes of the audit. Appeal brief at 8.
2. The auditors noted that these percentages were "not in
accord with
Section 4504 of the State manual" and that if the Agency had
not
previously accepted the allocation system, "our recommended
adjustments
would have been far in excess of those presently proposed."
"Report on
Audit of Administrative costs Under Section 1864 and 1902 of the
Social
Security Act," p. 11 (reproduced as State Att.
1.).