Department of Health and Human Services
DEPARTMENTAL APPEALS BOARD
Appellate Division
SUBJECT: LTE and IRS Drug Disallowances
DATE: October 26, 1992
Docket Nos. 90-101 90-139 90-145 90-158 90-161
90-166 90-183 90-191 90-196 90-219
90-224 91-45
A-92-2 A-92-143 A-92-230
Decision No. 1366
DECISION
Background
Fourteen states appealed determinations by the Health Care
Financing
Administration (HCFA) disallowing federal financial participation
(FFP)
claimed by the States under Title XIX (Medicaid) of the Social
Security
Act (Act). 1/ HCFA, on the basis of audits performed by the
Office of
the Inspector General (OIG) of the Department of Health and
Human
Services (DHHS), determined that the States had claimed FFP for
drug
products that were ineligible for Medicaid reimbursement because
the
drug products had been deemed either ineffective or "identical,
related,
or similar" to ineffective drugs.
These appeals arose over a two-year period and were assigned to either
one
of two Board Members, and thus.both have signed this decision as
Presiding
Board Members. Once the Board realized that the same drug
products were
the subject of disallowances for the different States, the
Board proposed
that these appeals be coordinated. The parties agreed to
this
proposal. These appeals were then stayed for approximately nine
months
while HCFA reexamined the disallowances. HCFA eventually decided
not to
withdraw any of the disallowances.
The Board has previously examined issues related to drug effectiveness
in
three decisions. In Illinois Dept. of Public Aid, DAB No. 667
(1985)
(Illinois), the Board decided that once the Food and Drug
Administration
(FDA) publishes a notice that a drug product is
less-than-effective
(LTE) in the Federal Register, states are put on notice
that Medicaid
reimbursement for the drug product ceases from the date of the
Federal
Register notice, plus an applicable grace period. The
Board
specifically rejected the argument that HCFA had the
sole
responsibility for identifying ineffective drug products which
were
ineligible for FFP:
[W]e will not condone a dangerously passive approach to
the
problem of ineffective drugs. Medicare and
Medicaid
beneficiaries' use of ineffective drugs can be hazardous,
and
the State clearly had an obligation to move as quickly as
it
reasonably could to stop reliance on these drugs.
Illinois at 8.
In Pennsylvania Dept. of Public Welfare, DAB No. 836 (1986), the
Board
found that states are also required to cease Medicaid payments for
other
drugs products not specifically named in a Federal Register notice,
but
which are "identical, related, or similar" to drugs named in the
Federal
Register (IRS drugs). The Board further found that, in the
particular
circumstances of that case, Pennsylvania's pharmacist would have
had no
meaningful difficulty, after consulting such national drug compendia
as
the National Drug Code Directory and "Facts and Comparisons,"
in
determining in a relatively short period of time the IRS drug status
of
the drug products at issue in that case.
In Louisiana Dept. of Health and Human Resources, DAB No. 1083
(1989)
(Louisiana), the Board explored further the process by which a state
is
able to determine whether a drug product is an IRS drug. The
Board
found that many IRS drugs can easily be identified from
compendia
commonly used by pharmacists, and that these drug.products
therefore
should be treated the same as FDA-specified LTE drugs. The
Board also
found, however, that a state was not responsible for the
identification
of all possible IRS drugs, noting that many drug products do
not appear
in the various compendia commonly used by pharmacists:
It is debatable as to what extraordinary lengths
Louisiana
should have gone to ferret out the information, but there is
no
HCFA rule or guideline establishing the extent of
Louisiana's
obligation . . . . In this context, Louisiana should not
be
held, with the benefit of hindsight, to be culpable for
not
doing the virtually impossible. The State should not be
denied
FFP for the continued use of such a drug until such time as
HCFA
or FDA informed the State that the drug was an IRS drug, or
the
drug appeared in compendia, or the State otherwise
discovered
(or reasonably should have) that the drug was an IRS
drug.
Louisiana at 16 (emphasis in original). The Board thus found that,
in
the absence of any publication by HCFA or the FDA of what drug
products
are considered IRS drugs, a reasonableness standard should be
applied in
determining a state's responsibility for the identification of an
IRS
drug.
In these appeals, disallowed claims for both LTE and IRS drugs are
at
issue. While the disallowed drug products vary from state to
state,
each State had claims disallowed for at least one, if not all, of
the
following three drug products: Bellergal, an alleged IRS
drug;
dipyridamole, an alleged LTE drug; and Naldecon and
Naldecon-related
drugs, all alleged IRS drugs. 2/ The States disputed
whether these drug
products were, in fact, LTE drugs or IRS drugs, and
whether the States
had sufficient information to make those determinations.
3/ .The States
also appealed disallowances for other alleged IRS drugs,
but did not
offer specific arguments for these drug products, suggesting
rather that
the whole process by which HCFA assigns the responsibility for
the
identification of IRS drugs to the States should be invalidated.
The
States also challenged the audit process from which the
disallowances
resulted.
Summary of Holdings
For the reasons discussed below, and as described more fully in
the
conclusion to this decision, we reach the following conclusions.
The
disallowances relating to Bellergal are sustained. The
disallowances
relating to dipyridamole are reversed. The disallowances
relating to
Naldecon are reversed, except for those claims after actual
notice to
the States of HCFA's position. We decline to invalidate
either the
process of State responsibility in the identification of and
termination
of payment for LTE and IRS drugs or the audit review process
which
resulted in these disallowances, except for certain
recalculations
explained below. Disallowances for other drugs about
which the States
made no specific arguments, except these challenges to the
system, are
sustained.
Overview of the Drug Review Process
The Federal Food, Drug, and Cosmetic Act of 1938 (FFDCA), 21 U.S.C. .
301
et seq., establishes a system of premarketing clearance for drugs
and
prohibits the introduction into commerce of any "new drug" unless a
new drug
application (NDA) filed with the FDA establishes the safety of
that
drug. Under the FFDCA, a "new drug" is one not generally
recognized by
qualified experts as safe for its intended use. In the
Drug Amendments
of 1962, Public Law 87-781, Congress amended the FFDCA
to direct the FDA to
refuse approval of a NDA and to withdraw any
approval for NDAs if, after
notice and opportunity for a hearing, it
found a lack of evidence that the
drug product involved was effective as
well as safe for use under the
conditions prescribed, recommended, or
suggested in its labeling.
The FFDCA provides an exemption from the definition of a "new drug," if
a
product was subject to the 1906 Food and Drug Act and "if its
labeling
contained the same representations concerning the conditions of
its
use." .21 U.S.C. . 321(p)(1). The 1962 amendments perpetuated
this
exemption, in a provision commonly referred to as the
"grandfather"
clause, as follows:
In the case of any drug which on the day immediately
preceding
the enactment date, (A) was commercially used or sold in
the
United States, (B) was not a new drug as defined by
section
201(p) of the basic Act as then in force, and (C) was
not
covered by an effective application under section 505 of
that
Act, the amendments made by this Act shall not apply to
such
drug when intended solely for use under conditions
prescribed,
recommended, or suggested in labeling with respect to such
drug
on that day.
Section 107(c)(4) of Public Law 87-871.
Most of the new drug products that were introduced in the market
through
the new drug procedures from 1938 to 1962 were submitted to a
scientific
board to determine the drug product's effectiveness for all
conditions
in its labeling. Each sponsor of a drug product with a
1938-1962 NDA
was required to submit to the FDA evidence for its labeled
indications.
After 1962 no new drug was allowed to be marketed unless it
received
approval from the FDA for safety and effectiveness for its
labeled
indication through the NDA process.
The FDA's conclusions on a drug product's effectiveness in the
Drug
Efficacy Study Implementation (DESI) program are published in
the
Federal Register as DESI notices and Notices of Opportunity for
Hearing
(NOOH). Each DESI notice and NOOH is assigned a DESI number,
which
applies to all of the drug products covered by the notice. If the
drug
product is effective for all its labeled indications, then the
DESI
notice will state the marketing conditions for the drug product.
If the
drug product is effective for some labeled indications, the DESI
notice
will state the marketing conditions for those effective
labeled
indications. If the drug product lacks substantial evidence
of
effectiveness for all its labeled indications, a NOOH is published
in
the Federal Register. The NOOH gives notice to the sponsor of the
LTE
drug product, and any interested party, which includes the
manufacturer
or distributor of a drug product identical, related, or similar
to the
LTE drug, of an opportunity to request a hearing and submit data
to
demonstrate the drug product's effectiveness.
Publication of a NOOH in the Federal Register results in ineligibility
for
reimbursement under the Medicare and Medicaid programs for the LTE
drug and
IRS drugs. Denial.of FFP begins after a grace period of 30
days from
the date of the NOOH's publication in the Federal Register.
During that grace
period, a state is to notify providers that the
affected drug will no longer
be eligible for reimbursement under the
Medicare and Medicaid programs.
Statutory and Regulatory Provisions for Drug Reimbursement
under
Medicaid
States have the option of including the costs of prescription drugs
in
their Medicaid plans. Section 1905(a)(10) of the Act. All of
the
States appealing here have elected to include the costs of
prescription
drugs in their Medicaid plans. Section 1903(i)(5) of the
Act, however,
denies federal funding under Medicaid for drug products which
are not
eligible for Medicare payments under section 1862(c) of the
Act.
Section 1862(c) prohibits payment for --
(1) a drug product --
(A) which is described in section 107(c)(3) of the
Drug
Amendments of 1962,
(B) which may be dispensed only upon prescription,
(C) for which the Secretary [of DHHS] has
issued a
notice of an opportunity for a hearing . . . on
a
proposed order of the Secretary to withdraw
approval
of an application for such drug product . . .
because
the Secretary has determined that the drug is
less
than effective for all conditions of use
prescribed,
recommended, or suggested in its labeling,
and
(D) for which the Secretary has not determined
there
is a compelling justification for its medical
need;
and
(2) any other drug product --
(A) which is identical, related, or similar
[IRS] .
. . to a drug productdescribed
in
paragraph (1) and
(B) for which the Secretary has not determined
there
is a compelling justification for its medical need .
.
. ..The FDA regulations describe an IRS drug
product
as follows:
An identical, related or similar drug includes other
brands,
potencies, dosage forms, salts, and esters of the same
drug
moiety as well as of any drug moiety related in
chemical
structure or known pharmacological properties.
21 C.F.R. . 310.6(b)(1).
The regulations then describe the process for identifying an IRS drug:
Where experts qualified by scientific training and experience
to
evaluate the safety and effectiveness of drugs would
conclude
that the findings and conclusions, stated in a drug
efficacy
notice or notice of opportunity for hearing, that a drug
product
is a "new drug" or that there is a lack of evidence to show
that
a drug product is safe or effective are applicable to
an
identical, related, or similar drug product, such product
is
affected by the notice. A combination drug product containing
a
drug that is identical, related, or similar to a drug named in
a
notice may also be subject to the findings and conclusions in
a
notice that a drug product is a "new drug" or that there is
a
lack of evidence to show that a drug product is safe
or
effective.
21 C.F.R. . 310.6(b)(2).
The regulations define a "new drug" as follows:
Any marketed drug is a "new drug" if any labeling change
made
after October 9, 1962, recommends or suggests new conditions
of
use under which the drug is not generally recognized as safe
and
effective by qualified experts. Undisclosed or unreported
side
effects as well as the emergence of new knowledge
presenting
questions with respect to the safety or effectiveness of a
drug
may result in its becoming a "new drug" even though it
was
previously considered "not a new drug . . . ."
21 C.F.R. . 310.100(c)..Analysis of Issues Relating to Specific
Drug
Products
BELLERGAL
Bellergal and Bellergal-S are drug products manufactured by the
Sandoz
Pharmaceuticals Corporation (Sandoz). HCFA disallowed claims for
the
Bellergal drugs because, according to HCFA, Bellergal is IRS to two
drug
products named in a May 6, 1983 NOOH (48 Fed. Reg. 20,495) (DESI
597),
Belladenal and Donnatal. 4/ Belladenal, a drug product also made
by
Sandoz, and Donnatal have as their primary ingredients belladonna
and
phenobarbital. Bellergal is composed of the following
ingredients:
belladonna, phenobarbital, and ergotamine tartrate.
The States argued that any disallowance for Bellergal should be
reversed
for the following reasons:
-- it is questionable whether Bellergal
was in fact a "new
drug" subject to a DESI notice and subsequent IRS
status;
-- the presence of an additional
ingredient, ergotamine
tartrate, in Bellergal changes its use and
renders it
substantially different from the LTE drugs named in the May
6,
1983 NOOH; and
-- the States could not have nor should
have known that
Bellergal was an IRS drug.
We conclude that any dispute concerning the "new drug" status of
Bellergal
did not interfere with the States' ability to identify it as
an IRS, that the
additional ingredient did not prevent Bellergal from
being IRS to the drugs
in the 1983 NOOH, and that the States should have
identified Bellergal as
IRS.
A. The "New Drug" Status of Bellergal
The States contended that an ongoing dispute exists between the FDA
and
Sandoz as to whether Bellergal was subject to a NOOH, and that
this
dispute impeded their efforts to identify Bellergal as IRS.
Sandoz
began marketing Bellergal in 1934. Sandoz took the position
that
Bellergal was exempt from the DESI provisions under.the
"grandfather"
clause of the FFDCA. Sandoz maintained that Bellergal was
subject to
the Food and Drug Act of 1906 and therefore, under the
FFDCA's
"grandfather" clause, was not a "new drug" since the labeling
for
Bellergal contained the same representations as to conditions for use
as
it did under the 1906 Act. Sandoz accordingly never submitted a NDA
for
Bellergal under the FFDCA. Sandoz concluded that Bellergal,
being
exempt from the "new drug" category, cannot be considered IRS within
the
meaning of 21 C.F.R. . 310.6, and that accordingly Bellergal was
not
covered by any NOOH as part of the DESI review process. Illinois
Ex. 3.
We find the States' argument concerning the effect of the legal
dispute
between the FDA and Sandoz unpersuasive. There is no evidence
that the
States were paralyzed by the Sandoz-FDA dispute over the status
of
Bellergal, or indeed, that they were even aware of it, during the
time
period covered by these disallowances (1984 through 1988 for
States
disputing Bellergal disallowances). The earliest indication we
have
from the record before us that the States knew of the controversy
over
Bellergal was a March 31, 1989 letter from Sandoz to Illinois,
detailing
Sandoz's discussions with the FDA, which was apparently written
in
response to Illinois' inquiries after receipt of the audit results.
Id.
Thus, the States' argument that the Sandoz-FDA dispute justified
their
delays regarding Bellergal is patently untenable.
Furthermore, while this Board is not the proper forum to resolve
any
conflict between Sandoz and the FDA over Bellergal, we note that
the
record before us suggests that this dispute was, in fact, resolved
since
Sandoz failed to challenge a 1983 FDA finding that Bellergal does
not
qualify for the "grandfather" exemption. 5/.B. The Effect of
an
Additional Ingredient in Bellergal
The States also argued that, even if Bellergal is subject to the NDA
and
DESI process, the presence of an additional ingredient,
ergotamine
tartrate, removes Bellergal from the ambit of the 1983 NOOH.
The 1983
NOOH concerned a review of certain drugs containing an
anticholinergic
in fixed combination with a barbiturate, which were
classified as
lacking substantial evidence of effectiveness for
various
gastrointestinal disorders. Among the
anticholinergic/barbiturate
combination drug products listed in the NOOH were
Donnatal and
Belladenal, whose primary ingredients are belladonna and
phenobarbital.
The States argued that the different combination of ingredients
in
Bellergal radically changes its use and renders it
substantially
different from Donnatal and Belladenal, since the addition of
ergotamine
tartrate gives Bellergal a wider range of uses than
Donnatal. While
conceding that the ingredient with the greatest volume,
phenobarbital,
is the same in both the LTE drugs and Bellergal, the States
contended
that the critical factor in determining a drug's effectiveness is
the
labeled indications of usage. The States referred to the
Physicians'
Desk Reference, wherein the labeled uses of Donnatal and
Belladenal
relate only to the gastrointestinal system, while Bellergal is
shown as
therapeutic in the endocrine and cardiovascular systems, as well as
in
the gastrointestinal system..Being composed of three
ingredients,
Bellergal is considered a combination drug product.
Concerning
combination drugs, the regulations provide:
Two or more drugs may be combined in a single dosage form
when
each component makes a contribution to the claimed effects
and
the dosage of each component . . . is such that the
combination
is safe and effective . . . .
21 C.F.R. . 300.50(a) (emphasis added).
The regulations further provide:
A combination drug product containing a drug that is
identical,
related or similar to a drug named in a notice may also
be
subject to the findings and conclusions in a notice that a
drug
product is a "new drug" or that there is a lack of evidence
to
show that a drug product is safe or effective.
21 C.F.R. . 310.6(b)(2). Thus, a combination drug product may also
be
ineffective when the drug product is made up of two or more
ingredients,
where one or more of these ingredients is the subject of a NOOH
and is
the basis for finding the drug ineffective.
In order for a drug product to be considered an IRS drug, therefore,
its
ingredients do not have to coincide exactly with those of a LTE
drug.
If the ingredients were exactly the same, then the suspect drug would
be
"identical" to the LTE drug. It is sufficient, however, to confer
IRS
status on a drug product if that drug is "related or similar" to the
LTE
drug. Here, of the three ingredients present in Bellergal, two of
them
in combination, belladonna and phenobarbital, were classified as LTE
in
the 1983 NOOH. It is difficult to see how any drug would qualify
as
"similar" if the addition of a small amount of one other
ingredient
could be presumed to overcome the substantial identity of all the
other
ingredients. The States have not shown any reason why they could
not
identify Bellergal as IRS by tracing the two matching ingredients.
Nor
have the States provided any evidence that their experts determined
that
the belladonna/phenobarbital combination somehow contributes to
the
effectiveness of ergotamine tartrate, as each ingredient in
a
combination drug must. 6/.The regulations are explicit that
"each
component" of a combination drug must make a contribution to the
drug's
effectiveness. 21 C.F.R. . 300.50(a). The Court of Appeals
for the
Third Circuit has held that "a product with an additional
active
ingredient must be more effective for its intended purposes than
that
product would be without the extra ingredient." United States v.
225
Cartons, More or Less, of an Article or Drug, 871 F.2d 409, 416
(3rd
Cir. 1989).
It is not sufficient for the States to point to additional indications
for
which Bellergal is labeled, beyond those for which the NOOH found
belladonna
and phenobarbital to be ineffective. While a drug must be
found
ineffective for all its labeled indications to be LTE, a drug
which is
related or similar to an ineffective drug is not reimbursable
unless a
compelling medical need has been shown. No such compelling
need has
been alleged here, and none is likely to exist, even if we were
to address
the issue, since ergotamine tartrate is approved in other
forms.
Accordingly, we find that Bellergal is IRS to the LTE drugs
listed in the
1983 NOOH.
C. Notice of Bellergal's Status as an IRS Drug
The States' final argument concerning the Bellergal drugs was that
HCFA's
imposition of a disallowance amounted to a retroactive sanction.
Citing
Louisiana, the States contended that they could not have known
that Bellergal
was IRS until they received the audit findings. The
States argued that
a reference to drug compendia after the 1983 NOOH
would have shown only the
different usages for the drugs. The States
asserted, moreover, that
their experts concluded that Bellergal was not
an IRS drug. 7/
The Board held in Louisiana that a state should not be subject
to
disallowances for IRS drugs for the period of time between
publication
of the Federal Register notice and the point when the state knew
or
should have known that the drugs were IRS. Louisiana at 3.
Applying
this criteria to the facts of these appeals, we find that the
States
reasonably should have determined from the 1983.NOOH that Bellergal
was
IRS to the LTE drugs listed there. The near-identity of
Bellergal's
ingredients to those of the LTE drugs could have been ascertained
from
drug compendia, not merely the existence of additional indications
for
Bellergal. In the Physicians' Desk Reference (44th Ed., 1990)
(PDR),
Bellergal is indexed, along with Donnatal and Belladenal, under
the
active ingredient phenobarbital. PDR at 322. The descriptions
show
that Belladenal consists of alkaloids of belladonna (.25 mg)
and
phenobarbital (50 mg), while Bellergal-S consists of alkaloids
of
belladonna (.2 mg), phenobarbital (40 mg), and ergotamine tartrate
(.6
mg). Compare PDR at 1931 with PDR at 1932.
Unlike the situation with Naldecon, discussed below, we have no
evidence
that the States' pharmacists actually made prospective
determinations
that Bellergal was not IRS in their professional judgment,
that the
States were in widespread agreement about such a conclusion, or that
any
efforts were made to contact FDA or HCFA for guidance about
Bellergal
specifically if the State experts found the determination difficult
to
reach. Furthermore, even before us, the States have not developed
any
persuasive explanation of why Bellergal would not be considered IRS
to
Belladenal and Donnatal, given the strong similarity of
ingredients.
Under the reasonableness standard set forth in Louisiana, the
States
have not shown that it was unreasonable to expect them to
have
identified Bellergal as IRS from an examination of national
drug
compendia.
Accordingly, we find that the Bellergal drugs were IRS drugs, and
we
sustain HCFA's disallowance of claims for FFP for the Bellergal drugs.
DIPYRIDAMOLE
Dipyridamole is a generic drug product manufactured by a number
of
pharmaceutical companies and identical to a brand name product
marketed
under the name Persantine. During the period at issue, a
NOOH
classified dipyridamole as less-than-effective for its only
labeled
indication. Before issuing the NOOH, however, the FDA approved
a
supplement to Persantine's NDA, finding it effective for a
new
indication, which its manufacturer was permitted to substitute on
its
label. Because of a law providing for a three-year
marketing
exclusivity, the manufacturers of the generic products were
not
permitted to amend their labeling to show the new effective
indication
until the exclusivity period expired. Since the LTE status
of a drug is
weighed in relation to its labeled indications, the generics
were
classified LTE in the interim..HCFA argued that the issuance of a
NOOH
relating to dipyridamole automatically triggered termination of
payment,
and no other facts should be considered relevant. The States
contended
that retroactive reimbursement for dipyridamole was guaranteed
by
language in the preamble to HCFA's regulations asserting that
Medicaid
would retroactively reimburse for any drug for which a NOOH was
issued
but which was later determined to be effective. In any case, the
States
argued that, because of the unique situation presented here, the
Act
could not have intended to cut funding for a product already
found
effective, simply because another law (for purposes unrelated
to
Medicaid) prevented the effective use from being added to the
labeling.
We conclude that dipyridamole is eligible for retroactive
Medicaid
reimbursement, and the disallowances for generic dipyridamole
are
consequently reversed. The States' challenge to the
automatic
triggering of non-payment by the NOOH in the case of dipyridamole
is,
thus, moot.
A. The Regulatory History of Dipyridamole
The history of FDA actions relating to dipyridamole is
relatively
complex. In 1971 the FDA classified the generic drug
product
dipyridamole, including a brand name version known as Persantine,
as
possibly effective for long-term therapy of chronic angina pectoris.
36
Fed. Reg. 3,078 (Feb. 17, 1971). Later, in 1972, the FDA
exempted
coronary vasodilators (anti-anginal drugs), including
products
containing dipyridamole, from the time limits established for
completing
the DESI program. 37 Fed. Reg. 26,623 (Dec. 14, 1972).
This exemption
allowed 81 dipyridamole drug products to remain on the market
while
studies were conducted to determine their effectiveness. FDA
granted
the exemption because of the medical need for, and absence of,
effective
drugs for treatment and prevention of anginal attacks. All of
these
dipyridamole drug products had only one labeled indication,
for
long-term therapy of chronic angina pectoris.
In 1977 the FDA amended the exemption by announcing guidelines and
methods
for evaluating the bioavailability and effectiveness of
coronary
vasodilators, and describing conditions under which the drug
products
could be marketed while the studies were in progress. 42 Fed.
Reg.
43,127 (August 26, 1977). On February 17, 1978, the FDA announced
that,
in the interest of minimizing duplicate and costly clinical testing
of
essentially the same drug product, each manufacturer need not
clinically
test its own product for effectiveness as long as at.least one
other
manufacturer conducted appropriate tests on a product containing
the
same chemical entity in a similar dosage form. 43 Fed. Reg.
7,044. The
FDA later announced that this new policy was implemented
specifically
with regard to single-entity coronary vasodilators. 43
Fed. Reg. 41,282
(September 15, 1978).
In September 1981, Boehringer Ingelheim (Boehringer), the manufacturer
of
Persantine, submitted to the FDA the results of four clinical
studies
intended to demonstrate the effectiveness of Persantine in the
long-term
therapy of chronic angina pectoris. On January 15, 1987, the
FDA issued
a NOOH which reclassified dipyridamole as "lacking substantial
evidence
of effectiveness" for long term therapy of chronic angina
pectoris,
proposed to withdraw NDAs for the labeled indication of angina,
offered
an opportunity for a hearing on the proposal, and revoked the
temporary
exemption that had allowed drug products containing dipyridamole
to
remain on the market. 52 Fed. Reg. 1,663. FDA found that the
four
studies submitted by Boehringer did not "constitute substantial
evidence
that Persantine is effective for long term therapy of chronic
angina
pectoris." Id. at 1,667. 8/
Although not mentioned in the January 15, 1987 NOOH, approximately a
month
earlier (on December 22, 1986) the FDA had approved a supplement
to
Boehringer's NDA for Persantine. The supplement approved was for a
new
labeled indication "as an adjunct to coumarin anticoagulants in
the
prevention of postoperative thromboembolic complications of
cardiac
valve replacements." HCFA Ex. 31. The FDA concluded that
"adequate
information has been presented to demonstrate.that the drug product
is
safe and effective for use as recommended" in the new labeling.
Id.
The approved labeling for Persantine contained the new indication
only.
Boehringer received a three-year marketing exclusivity for
Persantine
for the new labeled indication under the Drug Price Competition
and
Patent Term Restoration Act of 1984, Public Law 98-417, which amended
21
U.S.C. . 355. See 53 Fed. Reg. 12,605 (April 17, 1987). During
these
three years, no other manufacturer could receive approval to label
its
drug product for the new indication. The net effect of the
FDA's
actions was to inform the public that dipyridamole was ineffective
for
one use, effective for another, and subject to a delay in correcting
the
generics' labeling to reflect that reality. In the meantime,
generic
dipyridamole products remained on the market.
After the expiration of the three-year period, other manufacturers
of
dipyridamole products submitted, and received approval from the FDA
for,
supplements to their abbreviated NDAs to reflect the new
labeled
indications. See HCFA Exs. 25 through 29. In the interim,
HCFA
approved claims of FFP for Persantine, but denied FFP for all
other
claims for dipyridamole submitted by the States, even though they
were
identical and bioequivalent to Persantine. 9/ However,
HCFA
acknowledged that the other products are now eligible for FFP
upon
obtaining approval to substitute the new indication on their
labels.
HCFA Br. at 58-69.
B. Retroactive Reimbursement for Dipyridamole
Since dipyridamole has ultimately been proven effective for a
new
indication which has been substituted for the prior labeled
indication,
and is now reimbursable, the commitment made by HCFA to reimburse
DESI
drugs later found effective appears applicable. In the preamble
to.the
Federal Register notice promulgating 42 C.F.R. . 441.25, codifying
the
prohibition of Medicaid payments for ineffective drugs, HCFA stated:
If any drug that is the subject of a NOOH is subsequently
proven
to be effective, the Federal government will reimburse
under
Medicare Part B and Medicaid claims that were denied during
the
period the NOOH was in effect because the drug was
determined
less than effective.
46 Fed. Reg. 48,550 at 48,551-52 (October 1, 1981). Further, the
House
report on the bill adding the restrictions on LTE drugs to the
Medicaid
Act anticipated that, if a drug is contained in a NOOH based on
the
Secretary's determination that it is "less than effective for
all
conditions of use," but is "subsequently proven to be effective,
then
reimbursement would be allowed." H.R. Rep. No. 158, 97th Cong.,
1st
Sess. 345. These statements would appear to guarantee
retroactive
reimbursement for dipyridamole, since it was "proven to be
effective"
and had claims denied during a period when a NOOH was "in
effect."
HCFA argued that retroactive reimbursement is nevertheless
unavailable,
because the preamble language was intended to be limited to
situations
where the NOOH is determined to be in error. HCFA Br.
59-60. The plain
language of the preamble is not so limited.
Neither the preamble nor
the legislative history limited the intended
retroactive funding to
drugs whose original indications were
vindicated. A NOOH is a proposed
action of the Secretary to withdraw
approval for a drug because of a
determination that it is LTE for all
"conditions of use . . . in its
labeling." The DESI process of which
the NOOH was one stage cannot be
said to be limited only to consideration of
the initially-sponsored
indications. The final actions which ultimately
flow from a NOOH may
include withdrawal of the drug from the market, hearings
on or
investigations of the effectiveness of the drug with a
determination
that some or all conditions of use on the label are effective,
or, as
here, altered labeling to reflect effective conditions of use.
10/ HCFA
did not.deny that once the challenge to the effectiveness of
the
dipyridamole drugs was resolved by the substitution of revised
labeling
reflecting their effective use, they became eligible for FFP.
HCFA Br.
at 58-59.
Dipyridamole was the subject of a NOOH which was in effect during
the
period for which claims were denied. During the DESI process of
which
the NOOH was a part, the FDA withdrew approval of some of the
products
when their manufacturers did not choose to seek hearings, while
others
remained on the market with pending hearing requests. 11/ 52
Fed. Reg.
11,753 (April 10, 1987). The.products listed in the
withdrawal notice
included Persantine, but the notice specified that it
applied to the
applications only "insofar as they pertain to the original
indication."
At least one manufacturer, and probably others, withdrew its
pending
hearing request only when it was able to substitute the new
labeling
indication, as a result of supplementing its original
application. The
NOOH process thus may fairly be said to have ended
only with the
approval of the supplemented applications. The NOOH may
not have been
proven wrong, but it is plainly no longer "in effect," since
many of the
generics (and presumably all of them ultimately) are approved for
a
labeled indication. Therefore, the Secretary has manifestly
not
"determined that the drug is less than effective for all conditions
of
use prescribed, recommended or suggested in its labeling" now that
the
labels are revised to show an effective use.
Nothing in our decision limits HCFA's ability prospectively to
restrict
its commitment to situations where the outcome for a drug subject to
a
NOOH is the vindication of at least one of the labeled
indications
originally listed in the NOOH, provided that notice of the policy
is
given to the States. However, it would be unfair to impose such
a
restriction on the States in this case when it was not articulated
by
HCFA in either the regulation or the preamble.
We note that the approval of new labeling for dipyridamole occurred
in
essence as a part of the DESI review process for the drug.
The
dipyridamole drugs involved here continued to be available on the
market
throughout the period, the new indication was known to be effective
even
before the NOOH was issued, and the same DESI process
ultimately
resulted in their approval. We might reach a different
conclusion if,
long after the DESI process has been completed and a drug
withdrawn from
the market, some new use were found. Medicaid could not
be held
responsible to fund a drug retroactively when no effective use was
known
at the time, simply because some later research uncovered one (and
such
situations should be rare, since an ineffective drug would normally
be
withdrawn from the market within a reasonable period of time). But
in
this case there is no reason to conclude that the drug was being
used
for ineffective purposes,.since the effective use was known
throughout
the period the NOOH was in effect. We conclude that, under
these
circumstances, HCFA's commitment to reimburse retroactively for
drugs
later proven to be effective should be carried out in the case
of
generic dipyridamole.
We see no reason why Medicaid should be denied the benefits of lower
price
generic drugs, which continued to be marketed legally absent FDA
withdrawal
action, by denying retroactive reimbursement for their use.
The medical
community was informed by the approval of Persantine's
labeling, with the
explicit FDA ruling that it was bioequivalent, that
an effective use existed,
and that equivalent generics were available.
12/ An examination of the
PDR, for example, would have informed any
physician that Persantine was
dipyridamole and was indicated for the new
usage in cardiac valve replacement
(at 691) and that a number of
manufacturers produced generic dipyridamole (at
310). A responsible
physician or pharmacist might well permit generic
substitution in light
of the enormous price differential. Under HCFA's
interpretation, any
patient or third-party payor could benefit from such cost
savings except
the ones that perhaps most need to, Medicaid and Medicare.
Our conclusion serves the purpose of permitting Medicaid and
Medicare
patients to benefit from an effective drug.at the most economical
cost,
which is clearly in accord with the purposes of the Act. See,
e.g.,
Section 1902(a)(30)(A) of the Act; 42 C.F.R. ..
447.331-447.333.
Unquestionably, the savings involved are substantial.
13/ One State put
the monthly cost of a prescription for Persantine (25
milligrams) per
patient at $14.85, the cost of the generic dipyridamole at
$1.26.
Oregon Ex. 7.C. Assuming a FFP rate of 50 percent, one
monthly
prescription of Persantine would cost both the federal government and
a
state $7.42 each, while for generic dipyridamole the cost would
be
$0.63, about one-twelfth the price of Persantine. 14/
We thus conclude that claims for dipyridamole are eligible for
retroactive
reimbursement. In light of this conclusion, we need not
address the
States' arguments that dipyridamole was not LTE for all
indications, despite
issuance of a NOOH, or that Medicaid reimbursement
should have been
permitted, even if it was LTE, because of compelling
medical justifications.
15/.
NALDECON
Naldecon and Naldecon-related drugs (Naldecon drugs) are combination
drug
products. 16/ HCFA disallowed claims for the Naldecon drugs
because,
according to HCFA, the Naldecon drugs are "identical, related,
or similar" to
Dimetapp Elixir and Dimetapp Extentabs (Dimetapp) found
to be LTE in a
December 23, 1983 NOOH. 48 Fed. Reg. 56,854 (DESI
11935). The
States argued that it was unreasonable to expect them to
identify these drugs
as IRS to Dimetapp based not on the specific
ingredients they contain but
only on the fact that they each contain
more than one decongestant or more
than one antihistamine, since the
NOOH did not make clear that this was the
reason that Dimetapp was LTE.
We conclude that the States were not given
sufficient information to
have identified the Naldecon drugs as IRS.
A. The Regulatory History Relating to Naldecon
The history behind the FDA's determination that Dimetapp was a LTE
drug
product is as follows. In 1972 the FDA found that Dimetapp
Extentabs
was possibly effective, but that Dimetapp Elixir in oral dosage
form
lacked substantial evidence of effectiveness as a fixed combination
drug
product. 37 Fed. Reg. 15,022 (July 27, 1972). In April 1973
the FDA
issued a NOOH proposing to withdraw approval of a NDA for
Dimetapp
Extentabs as lacking substantial evidence of effectiveness since no
data
was submitted pursuant to the July 1972 notice. 38 Fed. Reg.
10,168
(April 25, 1973). The FDA invited any interested party with an
IRS drug
product, not the subject of an approved NDA, to respond to the NOOH
and
participate in a hearing. Id. Any person who wished to
determine
whether a specific product was covered by the NOOH was advised to
write
to the FDA. Id.
On December 14, 1973, however, the FDA granted a temporary exemption
for
certain prescription oral drugs.offered for relief of cough,
cold,
allergy, and related symptoms. The exemption was granted because
of the
similarity between prescription drugs for these uses and over
the
counter (OTC) drugs for the same indications which were subject to
an
ongoing OTC drug review program. 38 Fed. Reg. 34,481-82. As a
result,
the sustained release version of Dimetapp was exempted from the
April
25, 1973 NOOH for an indeterminate amount of time.
Gerstenzang
Declaration, HCFA Ex. 1, . 35.
On September 9, 1976, the FDA published an advance notice of
proposed
rulemaking to establish a monograph by the Advisory Review Panel on
OTC
Cold, Cough, Allergy, Bronchodilator and Antiasthmatic Products.
41
Fed. Reg. 38,312. In this monograph the FDA proposed to
establish
conditions under which OTC cold, cough, allergy, bronchodilator
and
antiasthmatic drugs are generally recognized as safe and effective
and
not misbranded.
In the December 23, 1983 NOOH, the FDA announced that it was revoking
the
exemptions for Dimetapp Extentabs and Elixir and stated the
conditions for
the manufacturer of Dimetapp to market the Dimetapp
products in a
reformulated version. The formulation of Dimetapp ruled
LTE consisted
of the following ingredients: phenylephrine hydrochloride
(a
decongestant), phenypropanolamine hydrochloride (a decongestant),
and
brompheniramine maleate (an antihistamine).
The FDA explained:
Brompheniramine maleate and phenylpropanolamine
hydrochloride
were both considered to be safe and effective by the OTC
review
panel . . . . The OTC drug review panel . . . also
concluded
that combinations containing a nasal decongestant and
an
antihistamine, as in reformulated Dimetapp Extentabs and
Elixir,
are safe and effective (41 FR 38326).
48 Fed. Reg. 56,854 (December 23, 1983). The Federal Register
reference
is to a page from the proposed monograph which sets various
criteria for
Category I (effective) and Category II (ineffective) combination
drug
products. Among the combinations of ingredients classified as
Category
I are combinations containing an antihistamine and a nasal
decongestant.
The FDA later approved a reformulated version of Dimetapp
conforming to
Category I and containing one decongestant,
phenylpropanolamine
hydrochloride, and one antihistamine, brompheniramine
maleate. Id..On
July 19, 1985, the FDA withdrew marketing approval of
the original
three-ingredient formulation of Dimetapp Elixir and Extentabs as
lacking
substantial evidence of effectiveness, and stated:
Specifically, each of the two nasal decongestants contained
in
these combination products has not been shown to contribute
to
the effectiveness of the products. Reformulations of
the
products have been approved as safe and effective.
50 Fed. Reg. 29,484.
Naldecon is composed of the following ingredients: two
decongestants,
phenylephrine hydrochloride and phenylpropanolamine
hydrochloride; and
two antihistamines, chlorpheniramine maleate and
phenyltoloxamine
citrate. Thus, Naldecon has among its ingredients the
two decongestants
in combination that were in the formulation of Dimetapp
found
ineffective.
In the September 1989 revision to the State Medicaid Manual, HCFA for
the
first time published a list, described as not all-inclusive, of IRS
drug
products. The Naldecon products appear on this list.
Section
4370.1, Addendum D at D-2.
B. The States' Ability to Identify Naldecon as IRS
HCFA argued that, while Naldecon is not "identical" to Dimetapp,
it
nevertheless is "related" or "similar" to Dimetapp in that Dimetapp
and
Naldecon share "known pharmacological properties" as set forth in
21
C.F.R. . 310.6(b)(1) because they both contain two decongestants.
HCFA
contended that, if the States had been monitoring the Federal
Register
properly, they would have been alerted to the fact that the
FDA
considered any two decongestants in combination ineffective in a
drug
product. HCFA argued that, under the criteria set forth by the
Board in
Louisiana, the States could have then easily identified Naldecon as
an
IRS drug product by consulting such compendia as Facts and
Comparisons
or the Physicians' Desk Reference, where the ingredients for the
drug
products are listed. HCFA acknowledged that the States could not
have
identified which drug products were IRS to Dimetapp merely
by
pinpointing one or two of its specific ingredients and eliminating
every
cold remedy listed in the compendia references which contained
those
ingredients. HCFA Master Br. at 48. Nevertheless, HCFA
insisted, the
States' experts could have readily identified Naldecon as IRS
to
Dimetapp, by recognizing the basis of the LTE finding in the
Federal
Register.notices as the presence of any two decongestants and
then
consulting the compendia.
We do not share HCFA's view that the FDA's notices in the Federal
Register
were so unambiguous that the States should readily have been
able to
determine the basis for the FDA's finding that Dimetapp was
ineffective and
to extrapolate that finding through compendia to arrive
at the conclusion
that Naldecon was an IRS drug product. We find
nothing in the wording
of the December 23, 1983 NOOH that would, as HCFA
contends, lead necessarily
to the conclusion that the presence of more
than one decongestant in a drug
product makes that drug product LTE.
The NOOH stated that the
three-ingredient formulations of Dimetapp were
ineffective. The NOOH
stated that no clinical studies were submitted in
support of the
products. No further explanation was given.
The NOOH then stated that brompheniramine maleate and
phenylpropanolamine
hydrochloride were considered safe and effective by
the OTC drug review
panel, and that the OTC panel concluded that
combinations containing a nasal
decongestant and an antihistamine in the
reformulated Dimetapp products were
safe and effective. This latter
finding contained a reference to the
1976 proposed monograph. This,
according to HCFA, was adequate notice
to the States that the
combination of two decongestants in a drug product
makes that product
LTE. 17/
HCFA's position that the reference in the NOOH to the monograph would
have
explained the basis for the FDA's finding that Dimetapp was
ineffective is
not persuasive for several reasons. The monograph
referred to in the
.NOOH was an advance notice of proposed rulemaking.
18/ The stated
intent of that notice was "to stimulate discussion,
evaluation, and comment .
. . . before any decision is made on the
recommendations of the Panel."
41 Fed. Reg. 38,312. The notice also
proposed that the conditions in
the monograph for Category I drugs (safe
and effective) and Category II drugs
(ineffective) be effective,
respectively, 30 days and six months after the
date of the final
monograph in the Federal Register. 19/ Id.
Thus, the monograph did not
make any final determination regarding the
effectiveness of OTC cold
remedies.
Moreover, the specific page of the monograph cited in the NOOH, 41
Fed.
Reg. 38,326, declared that a Category I (effective) drug product
may
have a combination from each pharmacologic group, such as a
combination
"containing an antihistamine and a nasal decongestant." On
the same
page, however, a Category II (ineffective) drug product is
classified as
including "more than two active ingredients from the same
pharmacologic
group." (emphasis added.) These categories offered
little meaningful
guidance to the States for determining why the original
version of
Dimetapp was ineffective, since it arguably could have fallen
under the
effective category (having an antihistamine and a
decongestant, but not
more than two of either). Furthermore, a later
version of the monograph
abandoned any insistence that two ingredients from
the same group are
automatically ineffective. 20/.In support of its argument
that the
States should have been able to discern why Dimetapp was
ineffective,
HCFA also referred to another page of the monograph, 41 Fed.
Reg.
38,420, where FDA proposed a regulation. This proposed
regulation,
which was to be codified at 21 C.F.R. . 341.40(b), stated that
any
single antihistamine active ingredient could be combined with any
single
oral nasal decongestant active ingredient. It is not clear that
this
language also meant that no drug could, like Naldecon, contain
two
ingredients from one of those categories. Even if the language
were
clearer, the States cannot fairly be considered to have been put
on
notice of the FDA's interpretation by a proposed regulation which
never
has been promulgated in final form. Proposed regulations are,
by
definition, subject to revision or even abandonment by an agency, so
the
public cannot be expected to rely on them. See, e.g., Powell v.
Andrus,
631 F.2d 699, 702, n. 2 (10th Cir. 1980); Joyce Faye Hughey, DAB
No.
1221, at 5-6, 8-9 (1991).
HCFA's position regarding Naldecon might thus be summarized as
follows:
the States should have recognized Naldecon as IRS based on a NOOH
that
did not contain any explanation for the action taken on Dimetapp,
but
did refer to one page of a seven-year old, 122-page proposed
monograph
that contained contradictory information, accompanied by a
proposed
regulation that, as of today,.some 16 years later, has never
been
promulgated in final form.
The States would bear an unreasonable burden if they are unable
to
determine from the explanation in the NOOH itself the basis on which
to
classify drug products as IRS to the products found to be
ineffective.
If the States are to meet their responsibility to identify IRS
drug
products, they must be given the necessary information to carry out
that
task. That was not done in the case of Dimetapp and,
consequently,
Naldecon. 21/
HCFA's position is further undermined by the fact that only a small
number
of states made the determination that Naldecon was an IRS drug
product.
The regulations provide that "experts qualified by scientific
training and
experience" are to make the determination whether a drug
product is IRS to a
LTE drug. 21 C.F.R. . 310.6(b)(2). The States
contended that
their experts had reviewed the NOOH for Dimetapp and
concluded that Naldecon
was not IRS to Dimetapp. A survey conducted by
one State showed that of
34 states audited by OIG, only two states, Utah
and Iowa, had identified
Naldecon as an IRS drug product. New Jersey
Appeal File at 15-20.
Another State pointed out that none of four major
drug database companies had
identified Naldecon as an IRS drug product.
Indiana Appendix C at 5. If
the basis for the FDA's finding in the NOOH
that Dimetapp was ineffective
were as clear as HCFA maintains it was, it
is probable that more states, as
well as the drug database companies,
would have made determinations that
Naldecon was an IRS drug product.
HCFA contended that if states have questions concerning a particular
drug
product, the states are encouraged to make inquiries to the FDA as
to the
applicability of a NOOH to that drug product. 21 C.F.R.
.
310.6(b)(3). HCFA pointed out that Idaho made such a request
concerning
Naldecon in July 1986, and received a response, through the
HCFA
regional office, in February.1987 that Naldecon was considered IRS
to
Dimetapp. HCFA Oregon Ex. 2. This does not, however, address
the
situation where a state has consulted its expert and that expert
has
rendered a professional opinion that the drug product is not IRS.
As
the States have pointed out, no HCFA or FDA rule imposes an
obligation
upon a state to verify that the FDA agrees with the opinions of
the
state's own experts. We also note that nothing in the record
indicates
that, after receiving the FDA's opinion that Naldecon was an IRS
drug,
the regional office disseminated that information to all the states
in
the region rather than just to Idaho. Further, other States
reported
they made efforts to obtain information in this regard from the FDA
and
encountered considerable difficulty. See, e.g., Missouri Br. at
10;
Maryland Br. at 3 and 13.
The Board has described the process for the identification of a drug as
an
IRS drug product:
A NOOH lists a LTE drug by its trade name, active
ingredients,
dosage form/route, and manufacturer. With this
information a
trained individual would consult various publications
and
national drug compendia . . . . From reference to
these
materials, that individual could presumably locate all IRS
drugs
in the marketplace. The most important information in
this
process is the active ingredients listed in the NOOH. As a
HCFA
witness testified, "[T]he burden is to follow the
ingredients.
If you have the ingredients, you have the key to the
whole
thing."
Louisiana at 12.
Here, however, HCFA conceded that the States could not have
identified
Naldecon as an IRS drug from merely the ingredients of Dimetapp
listed
in the 1983 NOOH. HCFA Master Br. at 48. The basis for the
FDA's
finding that Dimetapp was ineffective was not any of
Dimetapp's
ingredients specifically, but the fact that Dimetapp contained
two
decongestants. 22/ The FDA, however, as we.have shown above, failed
to
articulate that basis in the NOOH. It is not enough that a drug
is
listed in compendia; the NOOH must provide ingredients, or
other
articulated criteria, by which an expert can reasonably trace its
IRS
status. Hence, the States' experts could not have reasonably
determined
that Naldecon was an IRS drug through the use of compendia.
In order for the States to have carried out their responsibilities
with
respect to Naldecon, they required adequate information beyond
that
provided in the NOOH. The FDA and HCFA did not provide the States
with
that information. Accordingly, we reverse the disallowances
for
Naldecon and Naldecon-related drug products, up to the time HCFA
put
each State on notice -- through either direct communication or
the
September 1989 revision of the State Medicaid Manual, plus
the
applicable grace period -- that those drug products were considered
IRS
drugs. 23/ This result conforms with our holding in Louisiana that
the
States "should not be held, with the benefit of hindsight, culpable
for
not doing the virtually impossible." At 16.
Challenges to The LTE/IRS System Generally
The States contended that their experience with Naldecon is
illustrative
of the difficulties the States currently confront in
implementing the
DESI program. See discussion in the section on the
Audit Process,
infra. The States argued that it is fundamentally unfair
for HCFA to
require the States to make determinations on the IRS status of
drug
products, and then for HCFA to use one state's determination to bind
all
the other states. The responsibility for identifying IRS drugs
should
fall on the FDA, according to the States, and HCFA should then have
an
affirmative obligation to notify the States of the drug products
which
are considered to be IRS drugs. Until such notice is given, the
States
argued, Medicaid reimbursement should be available for drug
products
later determined to be IRS. The States generally blamed the
.federal
agencies for poor administration and communication.
One State argued that a 1990 OIG Report recommending changes in how
HCFA
and the FDA handle LTE drugs demonstrates that the federal government
is
wholly responsible for any payments for IRS drugs. Indiana
Supplemental
Br. at 4-8; "DHHS' Enforcement of Regulations Prohibiting
Medicaid
Payments for Less-Than-Effective Drugs," OIG/OAS Report No.
A-03-8900220
(July 13, 1990). We do not agree that this report in any
way exempts
the States from their responsibility during the audit periods
for
identifying IRS drugs.
The Board has previously heard such general attacks on the
supposed
inequities of the current DESI system and has rejected them.
Whether
this system is the most efficient alternative is a policy issue which
we
do not address. 24/ The Medicaid program is a partnership between
the
federal government and the states. Each partner has
responsibilities in
seeing that the Medicaid program is managed in a
responsible, safe, and
effective manner, including the removal of unsafe and
ineffective drug
products. The Board "will not condone a dangerously
passive approach to
the problem of ineffective drugs." Illinois at
8. The States cannot
simply wait for HCFA to inform them of each LTE
and IRS drug without
risking delays which could be avoided by proactive
efforts to .monitor
the NOOHs and identify IRS drugs. 25/ In this
effort, good faith, as
claimed by the States, is not enough. Effective
actions are required.
Therefore, we are not reversing the disallowances for any other IRS
drugs
appealed by the States where the States did not offer arguments
why they were
unable to identify those specific drugs as IRS. 26/.One
State did make a
specific argument relating to Pediacof Syrup, which was
disallowed as IRS to
DESI Notices 6514 as well as the Dimetapp NOOH.
38 Fed. Reg. 4,006 (February
9, 1973) and 47 Fed. Reg. 22,604 (May 25,
1982). DESI notice 6514
expressly stated that drug combinations of
antihistamines and expectorants
are "irrational." 47 Fed. Reg. 22,605.
The State's pharmacist did not
contend that he had any difficulty
recognizing that Pediacof Syrup contained
a combination directly
addressed in that NOOH. Rather, he disagreed
with the scientific merits
of the FDA conclusion in the NOOH. While the
States must exercise some
judgment in determining which drugs are IRS to
those drugs covered in a
NOOH, they may not disregard a NOOH because they
disagree with it.
Therefore, the disallowances relating to Pediacof Syrup are
sustained.
27/.The Audit Process
In addition to challenging the disallowances for specific drug
products,
some of the States also called into question the audit process used
to
arrive at the disallowances. Initially, the OIG began a review in
1985
of drug claims from seven states aimed at ascertaining whether they
were
effectively halting claims for FFP for LTE and IRS drugs. 28/
This
review resulted in disallowances, which HCFA stated are not at
issue
here, and led to a national audit. HCFA Master Br. at 13.
Several
States objected to the methodology used in the national audit.
HCFA submitted a declaration from the official with lead
responsibility
for the national audit review, who detailed the process
summarized here.
See Marion Declaration. A list of 4300 LTE and IRS
drugs, identified by
NDCs, was compiled as of August 1988 by the OIG from the
initial review,
an FDA list, and lists from four states (Pennsylvania,
Indiana,
Virginia, and Louisiana). 29/ The list was then reviewed by
the FDA as
of July 1989 to verify that each drug was LTE or IRS based on
a
published NOOH and to cross-check the NDCs for each drug. The
FDA
removed over 600 drugs and the OIG eliminated 32 others due to
problems
with their NDCS, resulting in a verified list of 3583. This
list was
compared by computer to all drugs for which the States claimed
FFP
during the audit periods. 30/
The procedure then followed for each State (with three
exceptions
discussed below) began with the submission on magnetic tape of
the
State's paid claims and a determination by comparison with the
LTE/IRS
list of the amount of FFP attributable to such drugs. Then, a
random
sample of 200 paid LTE/IRS claims was examined manually by
comparing
them to the original copies to verify that the magnetic tape
was
accurate. The States were given an opportunity to comment on
draft
reports. Disallowance .letters were then issued reflecting
any
adjustments which were accepted based on the States' comments.
The three States which diverged from this audit procedure were New
Jersey,
Illinois, and Indiana. In New Jersey, the magnetic tape
submitted by
the State contained claims for drugs paid from State funds
as well as those
for which FFP was claimed. In order to determine for
what percentage of
the LTE/IRS claims FFP was claimed, the random sample
of 200 claims used to
check the accuracy of the magnetic tape was also
evaluated to ascertain that
195 were claimed for FFP and five were
State-only claims. The
disallowance was reduced by a projected amount
to eliminate the State-only
claims. In Illinois, the OIG found that the
State had a master listing
of LTE and IRS drugs, which the OIG compared
to the computer list used in the
other audits. This comparison showed
that eight drugs which the OIG
listed as IRS were omitted from Illinois'
master listing. The OIG then
used Illinois' periodic payment reports to
calculate expenditures and FFP
attributable to those eight drugs. 31/
Indiana maintained periodic payment
reports by NDCs for drugs claimed
for FFP against which the OIG ran its
computerized list and found claims
for 286 LTE and IRS drugs for which the
related FFP was disallowed. 32/
The audit process was complicated somewhat by the issuance of
our
Louisiana decision, after which the OIG removed from its
computer
program all NDCs not listed in the compendia or sources referenced
in
that decision, reducing the verified list to 2850, as of May 10,
1990.
The OIG reexamined the pending audits and issued final
disallowances
which did not include any amount attributable to the 733 NDCs
removed
from the list. However, some States already had received final
audit
reports and disallowances prior to that reexamination. HCFA
agreed to
have the OIG reexamine those States' disallowances and make
any
resulting reduction in the disallowances. HCFA Master Br. at
15. HCFA
should identify and notify each affected State when
the.reexamination is
completed and what if any reduction in the disallowance
resulted. 33/
If any State wishes to appeal the results of the reexamination
on this
issue only, it may do so within 30 days of receiving such notice
from
HCFA.
Some States argued that the random sample was used improperly
to
extrapolate the disallowances, because the auditors did not claim
to
have checked every "hard copy" claim. See, e.g., New York Br. at
14-18.
34/ However, the States provided the magnetic tapes to the OIG
as an
accurate record of the claims paid, so they can hardly insist that
the
use of a sample to verify the accuracy of the tapes against the
original
claims copies maintained by the States was insufficient to permit
HCFA
to rely on the magnetic tape data. Despite the apparent confusion
among
some States, HCFA clearly stated that the sample was not used
to
calculate the disallowance but only to doublecheck the accuracy of
the
magnetic tape. Accurate recordkeeping is a responsibility of the
States
under the Act. See sections 1902(a)(6) and (27). The
disallowance
amounts were based on the computer comparison of all paid claims
from
the audit period against the OIG's list of LTE/IRS drugs, not on the
use
of any statistical sample of claims. .Some States argued that
several
NDCs on the HCFA computer list have been reassigned to drugs which
are
not LTE or IRS, which casts doubt on the validity of the audit
results.
See, e.g., Georgia Br. at 21. HCFA acknowledged that NDCs are
sometimes
prematurely reassigned. HCFA argued, however, that the FDA
verified
that the NDCs were still assigned to the drugs at issue as of July
1989
and that all the audit periods (except for the second New
Jersey
disallowance) involved were prior to that date. Therefore,
HCFA
contended that "recycled" NDCs could not have affected the accuracy
of
the audits. Nevertheless, HCFA agreed to recalculate the
disallowance
of any State which demonstrates that a particular NDC for which
a claim
was paid was not the drug listed for that NDC on the computer listing
of
LTE/IRS drugs. HCFA Master Br. at 29. Any State with such
evidence
should submit it to HCFA within 30 days of receiving this
decision.
HCFA should issue a written response to this issue, along with any
other
recalculations HCFA may make in the disallowance amounts.
The States presented a variety of arguments against the essential
fairness
and overall reliability of the audit results. 35/ Some
States
complained that the computer list.developed by HCFA should have
been
provided to the States as a tool to modify their drug formularies
or
payment records, instead of employed retroactively to penalize them
in
an audit. See, e.g., Georgia Br. at 20; Maryland Br. at 8;
Indiana
Supp. Br. at 2. Further, the States argued that, if HCFA and
the FDA
could not maintain a complete list of LTE/IRS drugs and had to
compile
this list from various sources including some State listings, then it
is
unreasonable to expect any one State to keep a complete, current list
on
its own. See, e.g., Missouri Br. at 3 and 7. In addition, the
efforts
of various States led to differing conclusions, so that "no two
states
agreed on the status of all prescription products," despite good
faith
efforts. Id. at 10. Some States argued that, since each
State is
required to identify and stop payment on LTE/IRS drugs, the OIG
should
have deferred to each State's expert in determining which drugs were
not
reimbursable, absent some showing of negligent procedures by the
State.
See, e.g., Virginia Reply Br. at 8-9. Consequently, the States
argued
that it was unfair and improper for the OIG to compile a list
from
sources outside the State being reviewed, such as other States which
may
have reached different conclusions about particular drugs.
Certainly, compliance efforts would have been enhanced if the States
had
had ready access to some central listing and guidance in
resolving
difficult cases. As noted above, a number of States
represented that
inquiries to the FDA or HCFA to seek information to resolve
questions
about specific drugs met with slow or unresponsive results.
It is not
our role, however, to impose policy judgment on the operation of
the
LTE/IRS program, but rather to determine the enforceability of
the
particular disallowances before us. In so doing, we do not find
it
unreasonable for HCFA to approach the task of evaluating the
States'
compliance with LTE/IRS drug restrictions by seeking to compare
State
payment files against a master list of unreimbursable drugs. Nor
is it
unreasonable to use a variety of sources to compile as comprehensive
a
list as possible and then to seek FDA verification. 36/ Our
prior
decisions have made clear that the States have the
affirmative
responsibility to identify and restrict payment on LTE/IRS drugs,
using
their own expertise. It.is not enough for the States to argue
that
their job could have been made easier by more federal coordination
and
better communication; the States must nevertheless shoulder their
part
of the job. The States have not shown any basis to invalidate the
audit
as a whole as unreliable or improper, and we therefore decline to do
so.
Conclusion
For the reasons discussed above, we hold as follows:
We find that Bellergal was a readily identifiable IRS drug
and
we therefore sustain any disallowance for that drug product.
We find that dipyridamole is eligible for
retroactive
reimbursement, and we therefore reverse the disallowances
for
that drug product.
We find that it was unreasonable to expect the States
to
identify Naldecon and the Naldecon-related drug products as
IRS
to Dimetapp, and we therefore reverse the disallowances
for
those drugs. However, we sustain any disallowance for
the
period after issuance of the September 1989 State
Medicaid
Manual, which provided notice to the States that Naldecon
was
IRS. (The record only shows a later disallowance for New
Jersey
that would be affected.) We also sustain any disallowances
for
the original formulation of Dimetapp, which was clearly LTE
(see
fn. 21).
We decline to invalidate the drug payment termination
process
under which the States share the responsibility for the
denial
of Medicaid reimbursement for IRS drugs. Thus, we sustain
the
disallowances for any other drug product for which the
States
failed to make any specific arguments, as well as
the
disallowances for Pediacof Syrup and Meprogesic Tablets,
for
which we rejected the States' arguments. We reverse
the
disallowance for V.V.S. only if that drug product is the
drug
V.V.S. Vaginal Cream ruled by HCFA in 1991 not to be IRS.
Finally, we reject the States' challenges to the audit
review
process, except for certain recalculations, as follows:
-- HCFA is to notify any State which was audited against
an
earlier LTE/IRS drug list and should reexamine its
claims
against
the.
final, reduced list. Any State may
appeal to us on this issue,
within 30 days of its receipt of
written notice of the results of
HCFA's reexamination.
-- Any State with evidence that its
disallowance
included amounts for a NDC that was reassigned at
the
time to a drug not on the LTE/IRS drug list
should
submit its documentation to HCFA within 30 days of
this
decision. Any State may appeal HCFA's written
response
on this issue to us within 30 days of its receipt.
We direct HCFA to adjust the amounts of the disallowances
to
reflect the above conclusions. As a result of this
decision,
there will be numerous recalculations of the
disallowances. Any
State that believes itself adversely affected
by any
recalculation determination may appeal that determination to
the
Board within 30 days of that determination.
__________________________ M. Terry Johnson
__________________________ Donald F.
Garrett
Presiding Board Member
__________________________ Norval D.
(John)
Settle Presiding Board
Member.
ATTACHMENT A
Docket No. State Disallowed Amounts Appealed
90-101 Illinois Dept. of $127,715 Public Aid
90-139 Oregon Dept. of Human $75,754 Resources
90-145 Indiana Dept. of Public $870,277 Welfare
90-158 Iowa Dept. of Human $647,116 Services
90-161 Missouri Dept. of $555,240 Social Services
90-166 Maryland Dept. of Health $179,164 and Mental Hygiene
90-183
Virginia Dept. of $813,448
Medical Assistance
Services
90-191
North Carolina Dept. $816,131 37/ of
Human
Resources
90-196 New Jersey Dept. of $736,000 38/ Human Services
90-219 Pennsylvania Dept. of $1,297,963 Public Welfare
90-224 Colorado Dept. of $111,871 Social Services
91-45
New York State Dept. $647,081 of
Social
Services.A-92-2
Georgia Dept. of
$895,873 Medical
Assistance
A-92-143 Texas Dept. of Human $2,394,934 Services
A-92-230 New Jersey Dept.
of $803,718 Human
Services.
ATTACHMENT B
State
Appealed Disallowed Drug
Products
Illinois Bellergal, Naldecon
Oregon
Naldecon, Dipyridamole,
Pediacof
Syrup
Indiana
Bellergal,
Naldecon,
Dipyridamole,
Dimetapp
Iowa
Bellergal,
Naldecon, Naldecon-related (Entex, Rutuss), Dipyridamole,
Pediacof
Syrup
Missouri Dipyridamole
Maryland
Naldecon,
Naldecon-related (Entex, Rutuss, Naldeate
Syrup), Dipyridamole
Virginia
Naldecon,
Naldecon-related, Dipyridamole, Vaginal Sulfa, other IRS drugs
North Carolina
Naldecon, Naldecon-related, Dipyridamole,
Vaginal Sulfa, other IRS drugs
New
Jersey
Naldecon, Tedral, Dipyridamole,
other IRS drugs
Pennsylvania
Bellergal, Naldecon, Naldecon-related
(Tedral, Entex), Dipyridamole, Vaginal
Sulfa, Pediacof Syrup, other IRS
drugs
Colorado
Bellergal, Naldecon,
Naldecon-related (New-Decongestant, Tri-Phen-Chlor),
Dipyridamole
New York Dipyridamole
Georgia
Naldecon,
Naldecon-related
Texas
Naldecon,
Naldecon-related, Dipyridamole, other IRS drugs.
ATTACHMENT
C
AUDIT PERIODS
Illinois
1/1/84-9/30/88 ACN
A-05-89-00102
Oregon
10/1/87-12/31/88 ACN
A-10-90-00001
Indiana
1/1/84-7/31/88 ACN
A-05-89-00081
Iowa
1/1/85-12/31/88
ACN A-07-87-00200
Missouri
1/1/85-1/31/89 ACN
A-07-89-00199
Maryland
1/1/87-6/30/88 ACN
A-03-89-00223
Virginia
4/1/84-6/30/88 ACN
A-03-89-00221
North Carolina 1/1/86-8/31/89 ACN A-04-90-02000
New
Jersey
1/1/87-6/30/88 ACN A-02-89-01021
7/1/88-9/30/90
Pennsylvania 6/1/84-3/31/88 ACN A-03-89-00602
Colorado
1/1/85-12/31/88 ACN
A-08-89-00232
New
York
1/1/87-6/30/88 ACN
A-03-89-00229
Georgia
9/1/86-6/30/88 ACN
A-03-89-00225
Texas
1/1/86-2/28/89 ACN
A-06-90-00101
.1. See Attachment A to this
decision for a listing, by docket number,
of
the States and the amounts of
their
respective appealed
disallowances.
2. See Attachment B of this decision for a listing, by state, of
the
disallowed drug products being appealed.
3. Each State submitted its own brief and appeal file, with the
later
opportunity to incorporate arguments made by other States. In
addition
to filing an individual response to each state brief, HCFA
also
submitted a Master Brief (Br.) applicable to all the arguments made
by
the States. When we refer to an argument offered by the States, it
is
possible that the particular argument was made by only one or some
of
the States. We will identify an exhibit (Ex.) by the name of
the
particular State that submitted it.
4. Sandoz manufactures two products, Bellergal and Bellergal-S,
which
contain the same active ingredients in different dosages and which
do
not differ in any way material here. We refer to both when we use
the
term "Bellergal."
5. The controversy between Sandoz and the FDA over Bellergal
concerned
a 1977 FDA determination that the Bellergal fell within the "new
drug"
category because Bellergal failed to meet the criteria for
"grandfather"
status. The FDA explained that Bellergal's labeling had
changed, with
some indications dropped from the current labeling and some
new
indications added to the current labeling not present in the
pre-1938
labeling. Illinois Ex. 4, at 2. Citing United States v.
Allan Drug
Corp., 357 F.2d 713 (10th Cir. 1966), cert. denied, 385 U.S. 899
(1965),
and 21 C.F.R. . 310.3(h), the FDA declared that the deletion of
old
claims and the addition of new claims for a drug product destroyed
its
"grandfather" status. Id.
Determinations of whether drugs are "new drugs" or are exempt under
the
"grandfather" clause fall within the authority of the FDA, subject
to
district court review. See Weinberger v. Hynson, Westcott &
Dunning,
412 U.S. 609, 624 (1973). Significantly, Sandoz never
challenged in
district court the FDA's determination that Bellergal did not
qualify
for the "grandfather" exemption. Additionally, the 1983 NOOH
that
listed Donnatal and Belladenal as LTE drugs expressly stated that
it
"encompasses all issues relating to the legal status of the
drug
products subject to it (including identical, related or similar
drug
products . . .) e.g., any contention that any such product is not a
new
drug . . ." 48 Fed. Reg. 20,495, 20,501. The NOOH then warned
that the
failure to request a hearing "constitutes a waiver of any
contentions
concerning the legal status of any such drug product."
Id. Sandoz made
no request for a hearing. Therefore, the FDA's
determination regarding
Bellergal appears final, thus undercutting the
States' position.
6. The FDA found ergotamine tartrate effective for prevention
of
vascular headaches, both alone as an aerosol and in combination
with
caffeine in tablet form. 37 Fed. Reg. 15,032 (July 27, 1975)
(DESI
5929).
7. However, two States which made specific arguments on Bellergal
are
those in which, HCFA asserted, the States' pharmacists admitted
during
their audits that they did not make any attempt to identify IRS
drugs,
since they believed it was the responsibility of the FDA. See
fns. 31
and 32 infra.
8. This January 15, 1987 NOOH specifically named 62
dipyridamole
products. A February 23, 1987 NOOH added 16 dipyridamole
products. 52
Fed. Reg. 5,501. A May 8, 1987 NOOH further added
three more
dipyridamole products. 52 Fed. Reg. 17,477. The States
argued that, in
the event the Board should find that dipyridamole was a LTE
drug, any
disallowances for drug products named in the February and May
NOOHs
should be measured from these later NOOHs, plus the applicable
grace
periods, rather than from the January NOOH. HCFA responded that
the
later NOOHs specifically referred to the January 15, 1987 NOOH,
and,
therefore, the January NOOH applied to all dipyridamole products,
and
that in any the other generics would certainly have been IRS to those
in
the first NOOH. In view of our conclusion that the States' claims
for
dipyridamole should be retroactively reimbursed, we need not
address
this matter.
9. We note that all but one of the States whose claims for
dipyridamole
were denied had disallowances for audit periods that began prior
to the
January 1987 NOOH that announced dipyridamole's LTE status. We
accept
HCFA's assertion that the disallowances were calculated to include
only
those claims for dipyridamole that were submitted after the
publication
of the NOOH and applicable grace period, and acknowledge that
HCFA
expressed willingness to recalculate the disallowances of any State
that
demonstrated otherwise. HCFA Br. at 62, n.35. However, this
issue is
moot, in light of the determination that all the claims for
dipyridamole
are eligible for retroactive reimbursement.
10. All the parties pointed to the uniqueness of
dipyridamole's
situation, which arose not only from the substitution of a
new
indication but the grant of a marketing exclusivity period, which
did
not prevent the generics from being sold as identical to Persantine
but
did prevent them from promoting their drugs for the new use.
The
confusion which resulted from the FDA's actions is demonstrated by
an
article in the March 20, 1987 newsletter of the American
Pharmaceutical
Association, Pharmacy Weekly (Maryland Ex. 14), which noted
that it was
unclear whether the FDA's action precluded Medicaid payment for
generic
dipyridamole and quoted a HCFA employee saying, "We have no
clear
answers. We are asking for our General Counsel to help us set a
policy
for this." Id. at 2.
The record before us shows that at least one State specifically
submitted
a written inquiry to HCFA regarding the implications of the
January 1987 NOOH
regarding Medicaid reimbursement of dipyridamole drug
products.
Maryland Ex. 15. In July 1987 HCFA responded that FFP would
be
available for Persantine, but not for other dipyridamole products.
Maryland
Ex. 17. There is no indication, however, that HCFA
communicated this
information to any of the other States. In fact, one
State alleged that
a HCFA representative specifically told two State
employees that FFP would
not be denied for dipyridamole even after the
January 1987 NOOH. Texas
Exs. A and B. Texas argued that, on the basis
of this representation,
HCFA should be estopped from taking any
disallowance action on claims for
dipyridamole. HCFA denied that its
representative gave any such advice
to Texas employees and argued that
Texas failed to establish that it met the
elements for asserting
estoppel against the federal government. See,
e.g., Heckler v.
Community Health Services, 467 U.S. 51 (1984), and Office of
Personnel
Management v. Richmond, 496 U.S. 414 (1990), reh. denied 59
U.S.L.W.
3137 (1990). Inasmuch as we are reversing the disallowances
for
dipyridamole, we see no need for determining whether an estoppel
defense
would be appropriate under these particular circumstances.
11. The parties have not raised, and we have not addressed,
any
distinction among dipyridamole claims based on whether the
manufacturers
pursued hearing requests or the dates that their abbreviated
NDAs were
approved. We presume that the products which were withdrawn
from the
market are not reflected in the claims, and our conclusion
on
retroactive reimbursement makes the date of each product's
approval
irrelevant.
12. Of course, it is possible that some physicians
prescribed
dipyridamole in either brand name or generic form for its
former
ineffective indication. Some claims may be for still other
indications
for which the parties have noted that the product has been
commonly used
by physicians. See, e.g., Oregon Ex. 7; New York Ex.
4. Such
"off-label" use is an accepted and lawful practice. See,
e.g., PDR at
Foreword. The methods used to handle claims for
prescription drugs
under Medicaid do not enable the States to distinguish the
use for which
a drug was prescribed. The anomaly which would result
from denying
retroactive reimbursement for the generic drugs here is that FFP
would
be paid for Persantine prescribed by a doctor for the treatment
of
chronic angina pectoris, an indication for which all brands
of
dipyridamole, including Persantine, have been found ineffective.
Yet
FFP would be denied (for a three-year period) for a prescription
of
dipyridamole filled by a generic, even though admittedly
chemically
identical and bioequivalent to Persantine, for purposes for which
it has
been ultimately proven effective and been approved for labeling.
13. One manufacturer of a dipyridamole product estimated in 1990
that
the nationwide market for dipyridamole was in excess of $110
million.
Missouri Ex. 5, at 2.
14. New Jersey determined that, on average, the brand name
Persantine,
in all strengths, cost 525% more than generic dipyridamole.
Indiana
Appendix B at 1. The total cost of generic dipyridamole during
New
Jersey's audit period was placed at $674,183, while the estimated
cost
if Persantine had been substituted was put at $3,651,070.
Indiana
Appendix B at 3.
15. The States also argued that FFP for dipyridamole claims should
be
allowed because there was "a compelling justification for its
medical
need" for dipyridamole as an antiplatelet drug and produced
articles
from medical journals to support this indication. See Oregon
Exs. 4, at
A, and 7, at D and E. Section 1862(c)(1)(D) of the Act
allows FFP if
the Secretary of DHHS has made a determination that such a
justification
exists. We agree with HCFA that the authority to make
such a
determination lies solely with the Secretary. The Board is not
the
proper forum to address this argument.
16. The terms "Naldecon-related" drugs, or simply Naldecon, are
used
here to include various drug products which allegedly are IRS
to
Dimetapp, in that these drugs have among their ingredients more than
one
decongestant or more than one antihistamine. Among the
Naldecon-related
drugs that are subject to disallowances here are the
following drug
products: Entex, RuTuss, Tedral New-Decongestant,
Tri-Phen-Chlor,
Naldelate, Naldelate Syrup, Nalgest, Naldagen, Naldec, Naldec
Syrup,
Quadrahist, Decongestabs, Par de Con, Sinucon, and other unnamed
drugs.
17. Rather than drawing the conclusion that it was the presence of
two
decongestants that made the old formulation of Dimetapp ineffective,
it
could also have been reasonably inferred that it was the presence of
a
particular ingredient, phenylephrine hydrochloride, that made the
old
formulation ineffective. The old, ineffective Dimetapp had that as
an
ingredient; the new, effective Dimetapp did not. If the States
had
reasonably inferred that phenylephrine hydrochloride had been the
reason
for finding original Dimetapp LTE, the States would have
erroneously
identified every cold product with that ingredient as its
only
decongestant as IRS. The FDA has never found, however,
that
phenylephrine hydrochloride is ineffective.
18. We further note that the reference was to a FDA notice
published
some seven years earlier. We question whether it is
reasonable to
require state personnel charged with monitoring NOOHs to
maintain an
indefinite backlog of Federal Registers in the event a future
NOOH might
obliquely contain a reference to one of them.
19. The FDA did later publish in the Federal Register as a notice
of
proposed rulemaking a "tentative final monograph" on the
recommendations
of the OTC review panel. This occurred on August 12,
1988. 53 Fed.
Reg. 30,522. The Board is unaware of any
publication of a final version
of the monograph.
20. In the "tentative final monograph," some 12 years later, the
FDA,
in response to public comments, wrote:
Category I active ingredients from the same therapeutic
category
that have the same mechanism should not ordinarily be
combined
unless there is some advantage over the single ingredient
in
terms of enhancing effectiveness, safety, patient acceptance,
or
quality of formulation. Thus, the [FDA]'s combination
policy
does not set limits on the number of ingredients from the
same
pharmacologic group that may be combined, provided data
are
presented to show the combination meets the necessary
criteria.
Combinations containing ingredients from the same
pharmacologic
group will be permitted if adequate data are presented to
the
agency.
* * *
The [FDA] agrees that no fixed limit need be placed upon
the
number of active ingredients in a combination product if it
can
be shown to be a rational, safe, and effective combination
with
a suitable target population.
53 Fed. Reg. 30,522, 30,535.
21. Indiana appealed a disallowance for claims for the
original
formulation of Dimetapp. Indiana argued that since Dimetapp
was later
found effective this disallowance should be reversed. This
argument
ignores the fact that the disapproved Dimetapp and the approved
Dimetapp
are two distinct drug products, with the former being specifically
found
LTE. While we are reversing the disallowances for Naldecon
and
Naldecon-related drug products, we sustain any disallowances the
States
may have received for the original formulation of Dimetapp.
22. The States also argued that a significant difference in the
dosage
amounts in the decongestant ingredients in Dimetapp and Naldecon
removed
the possibility that Naldecon could be considered IRS to Dimetapp,
with
the dosages of two decongestants in Naldecon providing a
therapeutic
effect that Dimetapp's decongestants lacked. See, e.g.,
Virginia Ex. 22
and Oregon Ex. 7. Since the disallowances for Naldecon
are being
reversed, there is no need to consider this argument further.
23. The only State before us which will apparently be affected by
our
holding that the September 1989 Manual announcement constitutes
notice
that Naldecon is an IRS drug is New Jersey, which had a
second
disallowance issued for the audited period July 1, 1988
through
September 30, 1990.
24. Some of the problems that the States experienced in
the
identification of IRS drug products may be alleviated by
HCFA's
publication of updated lists of IRS drugs in the State Medicaid
Manual.
If HCFA continues to timely update this list, the difficulties that
the
States have experienced in removing IRS drug products from
their
reimbursement rolls in a timely fashion should be reduced. With
the
publication of IRS drugs in the Manual, the States have
been
definitively put on notice of what drug products are considered
IRS
drugs by HCFA and hence are ineligible for FFP. The publication of
a
list of IRS drugs in the State Medicaid Manual does not,
however,
relieve the States of their continued responsibility to monitor
the
Federal Register for NOOHs and to act expeditiously on their own,
as
reasonably as can be expected, in removing drug products IRS to
LTE
drugs listed in the Federal Register.
25. In this regard, it is noteworthy that the States have
presented
arguments articulating difficulty identifying only relatively
few
allegedly IRS drugs, out of the hundreds of products listed as IRS
in
the final audit list. If the process of identifying IRS drugs
were
generally as difficult as the States argued here, it is reasonable
to
expect such arguments to have been made for far more drug
products.
Thus, we conclude, as we did in Pennsylvania, that the States are
able
to make IRS determinations in most cases without great difficulty.
26. Two States argued that HCFA issued disallowances for some
drug
products that were manufactured by small companies and therefore did
not
appear in national drug compendia. The States argued that,
under
Louisiana, disallowances should not be taken for such IRS drugs that
are
difficult to identify. North Carolina named three such drugs, UAD
Forte
Lotion, Bantuss HC Syrup, and Bantuss C Expectorant. Virginia
claimed
the following drugs were also unlisted in compendia: Tuss-Ade
TD Caps,
Bionade C-modified, Vertab, Unituss, Naldec, Quadrahist, Vaginal
Sulfa,
Isolate Cpd, Nafazair, ParDeCon, Oratuss, Naldegen, and Sinocon.
In addition to its assertion that its audit list was already reduced
to
exclude drugs not listed in compendia, HCFA responded that these
drug
products did, in fact, appear in national drug compendia.
Marion
Declaration, . 19. The States did not deny
this assertion in their
reply briefs, and we therefore decline to pursue the
issue of these
drugs any further. We do note, however, that many of
these drug
products are "Naldecon-related" drugs whose disallowances we
have
reversed on other grounds.
One State also argued that its disallowances for miscellaneous
drugs
should be reversed because HCFA's delays in this case prevented it
from
making effective arguments on these drugs since the State's
pharmacist
left State employment in the interim. New Jersey Br. at
4-5. This
State, however, failed to show why it could not have obtained
access to
any other expert assistance, if needed. We find no prejudice
to New
Jersey and sustain the disallowances for those drugs as to which
no
arguments were raised.
27. One State mentioned two drugs as examples of the confusing
nature
of the DESI process but did not effectively articulate why it could
not
reasonably have identified them as IRS drugs. North Carolina Br. at
6.
Thus, the State complained that it did not understand why one
sulfonmide
(V.V.S.) is IRS to the LTE drug called AVC, while another drug
product
(Sultrin Triple Sulfa) is not. Id. The State, however,
provided no
information about the ingredients, dosages, or indications of
these
drugs. We note, however, that in an August 1991 State Medicaid
Manual
transmittal HCFA announced that as of January 24, 1986, due to
product
reformulation, V.V.S. Vaginal Cream was no longer an IRS drug.
Due to
lack of information provided by the State, we are unable to
determine
whether this V.V.S. Vaginal Cream is the drug product V.V.S. of
which
the State complained. If it is, HCFA should adjust its
disallowance
accordingly.
The State also asserted that a generic product (Meprogesic Tablets)
was
declared IRS, while the corresponding brand name product (Equagesic)
was
"removed form the DESI list in June 1988." Id. The State
provided no
evidence to support this statement, and Equagesic still appears
as a LTE
drug on the State Medicaid Manual lists in 1989 and 1991. We
therefore
decline to reverse disallowances based only on these
unsubstantiated or
erroneous assertions by the State.
28. Those states were Pennsylvania, Maryland, Maine,
Virginia,
Connecticut, Illinois, and Massachusetts.
29. The NDC is supposed to be unique to a specific product and is
not
supposed to be reassigned until at least five years after a product
is
discontinued. Gerstenzang Declaration, HCFA Ex. 1, at . 31-b.
30. See Attachment C of this decision for a listing of each
State's
audit period.
31. HCFA also alleged that Illinois' drug expert stated during
the
audit that he did not attempt to review for similar and related
drugs,
since he believed that the FDA should make that determination
first.
HCFA Illinois Br. at 7, and Ex. 3.
32. HCFA stated that the State pharmacist admitted during the
audit
that he did not monitor the Federal Register for LTE/IRS drug
notices.
HCFA Indiana Br. at 2.
33. In particular, HCFA should notify Virginia as to whether its
audit
was based on the final list or whether it falls into the
category
affected here. See Virginia Reply Br. at 10. Virginia
objected to this
procedure on the basis that the OIG may decline to reaudit
it and that
HCFA cannot be relied on to perform any recalculation.
Therefore,
Virginia argued that the audit should be invalidated as
unreliable. We
decline to take this step. The discrepancy between
the FDA-verified
list and the final HCFA list resulted largely from HCFA's
adjustment to
reflect our decision in Louisiana, rather than from the lack
of
forthrightness or sloppy administration attributed to HCFA by
Virginia.
See Virginia Reply Br. at 9-12. The States have not
demonstrated any
inherent unreliability in the audits, and HCFA has offered
to
recalculate the disallowances to reflect the reduced number of drugs
on
the final list. We conclude that this approach is preferable
to
remanding these appeals to begin again the entire lengthy audit
process.
34. New York also argued that the State's paid claims history file
used
in the audit included claims from outside the audit period. New
York
Reply Br. at 9. Since New York's appeal involved only
dipyridamole,
which has been held eligible for FFP, we find this issue
moot.
35. Virginia challenged the validity of the audit based
discrepancies
between a list of LTE/IRS drugs on a computer diskette, which
the State
obtained from HCFA, and HCFA's audit program as to some of the
drugs
included and as to the NOOH dates listed. Also, Virginia argued
that
its State drug claim files showed different amounts of FFP
received
during the audit period. However, the LTE/IRS list used by
Virginia to
evaluate the reliability of the master audit list is described by
the
State only as having "presumably matched the list published by HCFA
in
the September 1989 State Medicaid Manual." Virginia Reply Br. at
11,
n.10. The audit list was derived from other sources, while HCFA
never
contended that the list in the Manual was comprehensive. See,
e.g.,
State Medicaid Manual, . 4370.3 (April 1983) (acknowledges that IRS
list
"may not include all drugs which are affected"); . 4370 (September
1989)
(declares that "this list is not all-inclusive"). Therefore,
the
existence of discrepancies between the two lists does not
demonstrate
that the audit list was unreliable. Further, the difference
in the
total FFP shown by state records and the OIG audit report is not
a
sufficient basis to conclude that the audit report was not
reliable,
since Virginia has not eliminated other possible reasons,
including
errors in calculation or unreliability of State records.
36. Our conclusion regarding Naldecon makes clear, however, that
the
States may not be responsible where their experts cannot
reasonably
identify a drug as IRS, simply because another State's expert has
done
so.
37. HCFA suggested that North Carolina's return of $325,095 left
only
$491,036 in dispute. HCFA North Carolina Br. at 2-3. North
Carolina,
however, reiterated that it continued to press its appeal of the
total
disallowance. North Carolina Reply Br. at 2-3.
38. In the course of the appeal, HCFA reduced New Jersey's
disallowance
by $5,510.50 for OTC