Department of Health and Human Services
DEPARTMENTAL APPEALS BOARD
Appellate Division
SUBJECT: Arkansas Department of Human Services
DATE: August 22, 1991
Docket No. 90-119
Decision No. 1273
DECISION
The Arkansas Department of Human Services (State) appealed a
determination
by the Health Care Financing Administration (HCFA)
disallowing federal
financial participation (FFP) claimed by Arkansas
under Title XIX of the
Social Security Act (Act) for drugs provided to
Medicaid recipients from
April 1 through July 31, 1989. The State had
paid for the drugs based
on use of the "average wholesale price" (AWP)
plus the dispensing fee
established by the Medicaid State Plan in effect
during the disallowance
period ($4.01). During Board proceedings, HCFA
agreed to reduce the
disallowance amount to $589,382. Essentially,
HCFA's current position
is that the AWP minus 10.5% of the AWP, plus the
$4.01 dispensing fee, acts
as an aggregate limit on the drug payments.
The State first argued that application of the AWP minus 10.5% to
the
period in question here was reasonable only if the increased
dispensing
fee adopted in a later amendment to the State Medicaid Plan is
also
applied. When HCFA responded that it was required to apply the
$4.01
dispensing fee, the State asserted, as it had previously, that the
AWP
represented the State's "best estimate of the price generally
and
currently paid for those drugs," as required by the
Medicaid
regulations. The State requested, and received, an opportunity
for an
evidentiary hearing to provide testimony the State said would show
that
the AWP constituted its "best estimate" during the disallowance
period.
For the reasons stated below, we conclude that the State could
not
reasonably consider the AWP to be its "best estimate" since the
State
was aware that pharmacists .generally paid less than that amount
and
since the State had no pertinent records to support a determination
that
the AWP represented the price generally and currently paid. We
further
conclude, however, that HCFA should reconsider its position on how
to
calculate the amount that should be disallowed as a result of
the
State's failure to timely develop a supportable estimate. Thus,
we
uphold the disallowance in principle, but remand to HCFA to
reconsider
the disallowance amount, in light of the factors outlined
below.
I. Background
A. Requirements for drug payments
Section 1902(a)(30)(A) of the Social Security Act requires Medicaid
state
plans to provide methods and procedures "to assure that payments
are
consistent with efficiency, economy, and quality of care."
Rules
published in July 1987, and in effect throughout the period
involved
here, control payments for drugs under Medicaid. 42 C.F.R
447.331 -
447.334. For some drugs, HCFA lists specific payment
limits. For
unlisted drugs and certain others, all called "other
drugs," a state's
payments --
. . . must not exceed in the aggregate, payment levels that
the
[state] agency has determined by applying the lower of the
--
(1) Estimated acquisition costs plus
reasonable
dispensing fees established by the [state] agency;
or
(2) Providers' usual and customary charges to
the
general public.
42 C.F.R. 447.331(b).
The "estimated acquisition cost" (EAC) is defined in 42 C.F.R. 447.301
as
the state's "best estimate of the price generally and currently paid"
by
providers for a drug.
The Medicaid regulations further require each state to make findings
and
assurances, in its state plan, that its Medicaid expenditures for
all
other drugs, in the aggregate, are in accordance with 42 C.F.R.
447.331,
and to maintain "data, mathematical or statistical
computations,
comparisons and any other pertinent records to support its
findings and
assurances." 42 C.F.R. 447.333.
. B. Drug pricing generally
There is a history of concern in the Medicaid program about
controlling
costs of drugs. One problem was states' use of the "average
wholesale
price" (AWP) of drugs as a measure of acquisition cost. The
"wholesale"
element could be quite misleading, because there were sales
promotions,
allowances and discounts which reduced the actual prices
pharmacists
paid below the AWP.
In the mid-1970's, HCFA had proposed to limit payment to
actual
acquisition costs, apparently to counter the practice of using the
AWP
in national drug pricing publications. Determining actual costs
was
burdensome, however, and the agency ultimately decided to specify use
of
estimated costs in its regulations (but HCFA explained in the
preamble
that it was specifically rejecting the suggested use of the AWP, on
the
basis that it frequently produced an inflated figure). 40 Fed.
Reg
34518 (1975). The estimate made by the state agency was to
be
consistent with "any drug price information furnished the program
agency
by the Department [HHS]." 45 C.F.R. 19.3(b) (1975).
An audit conducted in 1983 by the HHS Office of the Inspector
General
(OIG) in six states (including Arkansas) found that pharmacists'
drug
costs averaged about 16% below the AWP, and that in only 14 of
3,469
purchases examined did providers pay the AWP or higher (and then
for
extenuating reasons). HCFA Exhibit (Ex.) F, p. 4.
The current rules were adopted effective October 29, 1987; among
the
changes, the rules applied limits on an aggregate rather than
a
drug-specific basis. One purpose of this change was to give
some
flexibility to states to establish reimbursement methods. In
discussing
the requirements for findings and assurances and related
recordkeeping,
however, the preamble to the rule indicated that HCFA expected
the
states "to determine the estimated acquisition costs before
making
comparisons on the aggregate basis." 52 Fed. Reg. 28648, 28654
(July
31, 1987).
C. The Arkansas State Plan and proposed amendments
The Arkansas State Plan provision on drug payments which was in
effect
during the disallowance period (and was effective January 1,
1987)
states that the State will pay the lower of the provider's usual
and
customary charge or .the "cost of the drug plus a $4.01 dispensing
fee."
HCFA Ex. A. In practice, the State was paying the AWP plus
the
dispensing fee (if this total was lower than the usual and
customary
charge). 1/
On February 11, 1988, the State submitted an amendment to its State
Plan
(Transmittal No. 88-05) which would have defined estimated
acquisition
cost to be the AWP. HCFA asked for clarifying information
in a letter
dated April 12, 1988, citing to various studies as evidence that
"use of
the AWP is not the State's best estimate . . . ." HCFA Ex.
H.
In a response received June 22, 1988, the Director of the State's
Division
of Economic and Medical Services explained why the State
considered the
available information on drug prices to be outdated,
inconsistent, arbitrary,
and statistically incomplete. He expressed
concern, however, that "AWP
is an artificially high basis for
reimbursement." HCFA Ex. G, p.
5. He said the State was entering into
a bidding process for a pricing
survey to determine the actual
acquisition price and to modify the
reimbursement structure accordingly.
Id.
HCFA notified the State by letter dated September 20, 1988, that it
was
disapproving State Plan Transmittal No. 88-05, on the basis that
AWP
could not qualify as the State's best estimate under the regulation.
2/
After .receiving this notice, the State issued an invitation for
bids
for a survey and, on December 3, 1988, entered into a contract with
the
accounting firm of Myers and Stauffer to conduct a survey of
Arkansas
retail drug prices and of dispensing costs incurred by
pharmacies. Tr.,
pp. 61-62; 141-142.
HCFA advised the State, in a letter dated February 8, 1989, that, if
the
State had not implemented an acceptable plan amendment by April 1,
1989,
HCFA would begin to defer the federal share of prescribed
drug
expenditures. The State responded in a letter dated February 17,
1989,
informing HCFA that it had contracted with Myers and Stauffer to
conduct
a survey and that the State expected to receive the results by May
31,
1989 and to implement an EAC at "this time." State Ex. D.
In a letter dated March 31, 1989, HCFA stated:
Your expenditure report for the period April 1
through June 30 is
due in August and our actions at
that time will depend on the
actions which have been
taken by the State. As pointed out in your
letter, you should have the results of the survey by May 31
which
is well in advance of your expenditure report
due date. If you
take immediate action to base
your drug ingredient cost estimate on
verifiable
data from that survey and put an approvable
methodology
in place during the April-June quarter,
we will not include in our
review expenditures made
before the date the methodology is in
place.
State Ex. E.
The State encountered some delays in obtaining the survey
results
(discussed below), and did not receive the survey report until
about
June 14, 1989. Tr., pp. 69, 141-142. The report (which was
60 pages
plus attachments) found that the average discount from AWP in
Arkansas
was 13%. HCFA Ex. B. The State sought some
clarifications from Myers
and Stauffer, held meetings to discuss the survey
results, developed a
state plan proposal and received internal approval for
it, and submitted
the proposal to HCFA on July 21, 1989. Tr., pp. 23,
69-74. The
proposed plan amendment, Transmittal No. (TN) 89-24,
would
have.established the EAC at AWP minus 7% and raised the dispensing
fee
to $4.39 plus .095(EAC). State Ex. F. 3/
After engaging in discussions with HCFA, the State submitted a
revised
version of plan amendment TN 89-4, which defined the EAC as AWP
minus
10.5% and set a dispensing fee of $4.16 plus .093(EAC). State Ex.
H.
HCFA approved the revision in July, 1990, with an effective date
of
August 1, 1989. State Ex. G. 4/
Meanwhile, HCFA had issued this disallowance on May 2, 1990,
calculating
the disallowance amount using AWP minus 13%. Based on the
revised plan
amendment, HCFA reduced the disallowance amount (and also
settled a
related disallowance for a later period).
II. The parties' arguments
The State argued that it has always been considered the role of the
State
to determine how it would arrive at the best estimate of the cost
of the
drug. The State asserted that, up until a few years ago, 20
states had
used AWP. The State acknowledged that since that time there
has been a
move to become more accurate about the "real cost" of drugs.
Tr., p.
10. The State said, however, that it had been actively engaged
since
1987 in assessing what is the best estimate, but that there was no
scientific
study to assess the costs in Arkansas until the Myers and
Stauffer study was
completed in June 1989. The State argued that it was
generally accepted
and recognized nationwide that the AWP should be used
"until such time as
there was objective scientific information .to the
contrary." Tr., p.
12. The State asserted that, once it had such
information, it had acted
promptly to evaluate that information, to
propose a plan amendment, and to
implement the proposed amendment. The
State argued that it had the
burden only to develop what it considered
to be the best estimate and that it
had met that burden.
The State also argued that it did not have notice until HCFA rejected
that
plan amendment specifying AWP in September 1988 that HCFA would not
consider
an undiscounted AWP to be consistent with the regulatory
requirement.
The State said that it had informed HCFA of the delays in
the survey process,
arguing that even with the delays, the survey was
completed in less time than
such surveys normally take.
HCFA argued that the State's payment of the AWP did not comply either
with
the regulation or with the State Plan. HCFA said that the State
should
have known it could not pay AWP, in light of the OIG report
(which State
officials acknowledged having seen) and of the statements
about the AWP HCFA
had made. HCFA also cited the statement in the April
12, 1988 letter
from HCFA that "use of an unmodified AWP would make the
amendment violate
Section 1902(a)(30) of the Act and implementing
regulations at 42 C.F.R.
447.200." HCFA Ex. H. HCFA argued that,
although none of the
various drug price surveys and pricing data were
conclusive for Arkansas,
"they collectively show a persuasive pattern
that AWP is not the actual
selling price of drugs anywhere such surveys
were conducted." HCFA
post-hearing brief, p. 8. HCFA argued that the
State was unreasonable
in delaying until September 1988 to commence a
process to determine drug
prices in Arkansas.
HCFA further argued that it had acted reasonably in quantifying
the
disallowance by applying the AWP minus 10.5% since the State had
not
properly quantified what price was being paid for the drugs and
since
the AWP minus 10.5% was the amount ultimately agreed to based on
the
survey data, which was from 1988.
Below, we first address the question of whether the AWP qualified as
the
State's "best estimate of the price generally and currently paid"
during
the disallowance period. We then address whether HCFA
reasonably
calculated the amount of unallowable costs incurred as a result of
the
State's use of the AWP.
.III. Analysis
The key issue here is whether the AWP can reasonably be considered
the
State's best estimate during the period in question here. While
the
State's evidence demonstrates the difficulties of determining a
best
estimate, the State's evidence ultimately does not support the
State's
use of the AWP.
The regulations clearly required that the State use its best estimate
of
the amount generally and currently paid. The regulations
contemplated
an effort by the State to determine this amount in some
reasonable
manner which could be documented. As this Board said in
Oklahoma Dept.
of Human Services, DAB No. 1271 (1991), p. 8, developing a
"best
estimate" of an EAC clearly requires at least some minimal
assessment
and determination of actual drug costing practices.
Moreover, the State
was subject to the general requirement in section
1902(a)(30)(A) that
its payment methods be consistent with economy. In
spite of this
obligation, the State had absolutely no data to support use of
the AWP.
See Tr., pp. 63, 91. What the State's evidence did show is as
follows.
The unrebutted testimony presented by the State at the hearing shows
that
the inappropriateness of using the AWP was not as clear from the
State's
viewpoint in the mid-1980's, as HCFA asserted it should be.
Other states were
using the AWP, and the AWP compendia were readily
available, with weekly
updates which could be loaded onto the State's
computerized claims processing
system. Tr., pp. 22, 49. The OIG report
was not based on a valid
statistical sample, and thus did not provide a
reliable basis for determining
how much the AWP was being discounted in
Arkansas. Tr., pp. 49-50,
65.
The testimony further shows that the State encountered
administrative
difficulties in the process of contracting for a reliable
survey and
obtaining and evaluating the survey results. The invitation
for bid had
to be approved by HCFA, and the State had to follow its
contracting
procedures. Tr., p. 90. Delays were caused through
the need to obtain
information from the State's fiscal agent. Tr., pp.
24-25, 66.
Moreover, the State had to rely on the pharmacists to respond to
the
survey. Tr., p. 24. The task of analyzing the data was
apparently
enormous, but Myers and Stauffer nonetheless completed the survey
in
less time than it took for similar surveys in other states. Tr.,
pp.
146-148. .The testimony also shows that the question of
determining
what discount percentage should be applied is not as simple as
HCFA's
arguments suggested. The amount of the discount given
varies
considerably (for example, chain pharmacies might obtain a
larger
discount than independent pharmacies). Tr., pp. 123, 152; HCFA
Ex. B,
Appendix, p. A-67. This raises an issue of how a state can set
a
discount which will ensure accessibility to services for
Medicaid
recipients and avoid legal battles over its reimbursement to
providers.
Tr., pp. 70-71. This was a legitimate concern for the State
since a
survey by the Arkansas College of Pharmacy had shown that, if the
State
were to set the reimbursement level too low, many pharmacists would
drop
out of the Medicaid program. Tr., p. 125; see also Tr., p.
71. Also,
undisputed testimony shows that the "discounts" which the
wholesalers
were giving to the providers were not all cash discounts.
Rather, in
some instances the wholesalers were simply shifting costs which
had
previously been included in the drug prices to appear as a
separate
charge on the sales invoices. These costs included charges for
items or
services such as labels, bottles, and computer ordering
services. Tr.,
pp. 117-121.
This evidence is an insufficient basis for the State to prevail
here,
however. First, as discussed above, the State did not have any
data to
support use of the AWP, as required. Second, while the
testimony shows
that the State also did not have the data to support a
determination of
the exact amount of the discounted prices being paid in
Arkansas, the
evidence shows that the State could not reasonably think that
the full
amount of the AWP was being paid generally. 5/ The OIG audit
certainly
was reliable at least in concluding that the AWP was not being
paid.
The State Director acknowledged that he knew of the OIG audit in
1986
and was on notice through it of the discrepancy between AWP and
the
actual price paid. Tr., p. 56. The Director of the Division
of
Economic and Medical Services acknowledged that, based on the OIG
audit,
HCFA had encouraged the State to look at setting some rate
of
reimbursement less than AWP. Tr., p. 64.
.Moreover, the State's explanations of its delay in obtaining
reliable
data do not justify the initial delay caused through the State
proposing
in 1988 a plan amendment specifying AWP, nor do they justify the
full
time period it took to obtain and act on the survey results. One
State
witness speculated that the State did not act sooner to commission
a
survey because HCFA was considering a new regulation discounting AWP
by
10% and because a lawsuit had been filed by pharmacists who
challenged
HCFA's position. Tr., pp. 101-102. The record shows,
however, that he
was referring to the regulatory process which was concluded
in 1987 and
that the lawsuit was concluded in 1985. See, e.g., HCFA
post-hearing
brief, attachment. Thus, this does not explain why the
State acted in
early 1988 to propose a plan amendment specifying AWP, rather
than to
obtain reliable data on the prices generally and currently paid.
The record shows that the State was not justified in waiting to begin
the
process of contracting for a survey until after the State received
formal
disapproval of the plan amendment specifying AWP, in September
1988.
The State's own letter sent to HCFA in June 1988 shows that the
State was
aware of the need for a survey at least as of that date and
had assured HCFA
it would act expeditiously to commission a survey.
HCFA Ex. G, p. 5.
The State's testimony about the need to follow
contracting procedures was
vague and neither specifically nor fully
accounts for the three months
between the June letter and the time the
IFB was issued in September.
Tr., p. 90. Moreover, we do not see how
the State could have reasonably
thought that HCFA might approve the AWP
as EAC, given HCFA's April response
and the history of HCFA's position
on AWP.
The State also did not provide any explanation of why it took as long
as
it did after receiving the proposals to actually award the
contract.
The witness from Myers and Stauffer said that they predicted a
report
date of April 1989 based on an assumption that the award date would
be
only two weeks or so after they mailed the IFB back, but, instead,
a
little over a month elapsed. Tr., pp. 144-145. Like this delay,
some
of the delays once the contract was awarded, such as obtaining
needed
computer lists from the fiscal agent, would appear to be ultimately
the
State's responsibility. See Tr., pp. 24-25, 145. While these
delays
were not lengthy, they contributed to the State's failure to meet
the
deadline HCFA announced in its March 31, 1989 letter.
.Finally, we conclude that the State could not reasonably have relied
on
the mere fact that it informed HCFA about the delays in obtaining
the
results of the survey as a basis for thinking that HCFA would extend
the
deadline announced in HCFA's March 31, 1989 letter. The State did
not
allege that HCFA had specifically agreed to extend the deadline.
See
Tr., pp. 51-52, 68. At most, the HCFA letter can be viewed as
a
conditional agreement to waive a disallowance. The State's failure
to
meet the condition in the letter by submitting a plan amendment based
on
the survey by the end of the quarter specified means that the
State
could not reasonably expect HCFA to refrain from a disallowance
solely
based on HCFA's silence. While State officials indicated that
they may
not have understood that payments as of April 1, 1989 would
be
disallowed if the State did not act by the date specified in the
letter,
their understanding is irrelevant; the March 31, 1989 letter gave
clear
notice of HCFA's intention.
In sum, we conclude that the State could not reasonably consider the
AWP
to be its "best estimate of the amount generally and currently
paid;"
that the State had constructive notice in 1987 that it needed data
to
support such an estimate; and that the State actually knew that a
survey
was needed at least by June 1988. We further conclude that HCFA
was
reasonable in setting a deadline for the State and in imposing
a
disallowance when the State failed to meet that deadline.
On the other hand, the record establishes that, in determining
the
disallowance amount, HCFA applied the AWP minus 10.5%,
without
considering the other changes in the reimbursement method HCFA
agreed
to. If the State had implemented its new reimbursement methods
so they
were effective during the disallowance period (as HCFA said the
State
should have), the amount the State would have paid would have
differed
from the amount the State did pay (using AWP) by an amount less than
the
disallowance amount HCFA calculated. The increase in the dispensing
fee
would also have been reflected in the State's claims, since this
was
part of the amendment agreed to. See State Ex. G.
We think that HCFA should reconsider its position on how to calculate
the
disallowance amount, in light of the testimony presented at the
hearing and
the other considerations we outline below.
HCFA indicated that it could not approve the new State plan
provision
retroactive to April 1, 1989 because the State had not implemented
the
provision until August 1,.1989. HCFA thus thought that it was
precluded
from using any dispensing fee other than the $4.01 fee specified in
the
existing State plan.
While certain state plan provisions may not be effective
until
implemented, regulations which HCFA published in 1988 permit a
plan
amendment such as this one to be effective on "a date requested by
the
State if HCFA approves it." 42 C.F.R. 430.20(b)(2) (1988).
HCFA's own
witness said that HCFA had expected the State to apply an
amendment
based on the survey back to April 1, 1989, by retroactively
adjusting
the provider payments. Tr., p. 183. The State's
rationale for not
doing so was that it had to give timely notice to the
providers. See
Tr., p. 67. If HCFA is correct that the payments
(at least for FFP
purposes) could be adjusted retroactively, then we see no
reason for
denying the State's request for an April 1, 1989 effective date
for the
plan amendment.
Even if HCFA would not grant this approval, however, it is not clear
why
HCFA applied part of the new methodology, but not all of it.
The
regulations on which HCFA relied for the disallowance
establish
aggregate upper limits on what the State can pay. The
regulations use
as that aggregate upper limit the lower of EAC "plus
reasonable
dispensing fees" or usual and customary charges. Although
the State may
have actually paid only $4.01 as a dispensing fee, the State's
survey
(which as HCFA pointed out was based on 1988 costs) establishes that
the
amount of a reasonable dispensing fee for this period was the
higher
figure HCFA ultimately agreed to.
Moreover, HCFA viewed the term "cost" in the existing plan as
"ingredient
cost," entirely separate from the dispensing fee. The plan
merely says
"cost," however, and the State's testimony showed that the
wholesalers had
been including as drug costs some items that more
properly should have been
considered part of the dispensing fee. See
Tr., pp. 49-50. The
State's testimony shows that, if the State had not
increased the dispensing
fee at the same time as discounting from the
AWP, the State would have had to
discount by a lesser percentage than
10.5 in order to assure access to
services, particularly for recipients
in rural areas. As the State
pointed out, it must not only consider
efficiency and economy in setting
reimbursement rates in accordance with
section 1902(a)(30)(A) of the Act, but
must also consider quality of
care.
.HCFA's position was essentially that the State was required under
the
regulations to have implemented a new reimbursement system at least
by
the beginning of the disallowance period (April 1, 1989). 6/ Thus,
we
conclude that HCFA should consider determining the disallowance
amount
by calculating the difference between what the State did pay and
what
the State would have paid if it had implemented the new system by
April
1, 1989 (including application of the usual and customary charge
limit).
Conclusion
We uphold the disallowance in principle, but remand to HCFA to
reconsider
how to calculate the disallowance amount. If the State
disagrees with
HCFA's recalculations, it may return to the Board on that
limited issue,
within 30 days after receiving notice of HCFA's
recalculations.
_____________________________
Donald
F. Garrett
_____________________________
Norval
D. (John) Settle
_____________________________
Judith
A. Ballard Presiding
Board Member
.1. At the
hearing, HCFA indicated that it was in the
process of calculating a
reduction in the disallowance amount to reflect the
fact that the State
had not paid the AWP plus dispensing fee where this total
was greater
than the usual and customary charge. Transcript of hearing
(Tr.), p.
17. We assume that the $589,382 figure used in HCFA's
post-hearing
brief reflects this reduction, or, if it does not, HCFA will
make this
adjustment, as agreed.
2. The State requested a hearing on the plan approval, but
the
administrative law judge appointed as the hearing officer stayed
the
case pending appeal of the related disallowance. We note that a
court
has held that HCFA may properly deny a Medicaid state plan
amendment
which proposes to set a state's EAC at AWP in the absence of a
showing
that AWP is in fact that state's "best estimate" of the price
generally
and currently paid. Louisiana v. U.S. Dept. of Health and
Human
Services, 905 F.2d 877 (5th Cir. 1990).
3. The State presented testimony explaining why it had not
adopted
the average discount from the AWP (13%) reported in the survey
report.
Essentially, the State was concerned that since about 30% of
the
pharmacies were paying more than the average amount, these
pharmacies
would drop out of the program if the average discount were adopted
as
the EAC and that this would reduce accessibility of services.
See Tr.,
pp. 33-35, 70; HCFA Ex. B, Appendix, p. A-67.
4. HCFA first said that it chose this effective date
because this was
the date the State requested in its transmittal. When
the State asked
if HCFA could nonetheless make the amendment retroactive HCFA
declined,
apparently on the basis that the State had not implemented the
amendment
until August 1, 1989. See our discussion of this below.
5. Indeed, even the Arkansas Pharmacy Association
representative said
that nearly all pharmacies were receiving at least a 2%
cash discount.
Tr., p. 139.
6. Thus, this case is distinguishable from the situation
where a
state is making an optional change in its reimbursement system in
order
to enhance federal funding and seeks to make that change
retroactive,
without following plan amendment