Research · 6 min read
What Decides Whether Part B or Part D Pays for a Drug
Medicare's drug benefit is defined by subtraction. A drug that Part A or Part B would pay for is, by regulation, not a Part D drug at all — and that single sentence explains most of the confusion about which card to hand over.
Key takeaways
- The Part D drug definition excludes any drug for which payment is available under Part A or Part B, so Part D begins where Part B's authority to pay ends.
- That exclusion applies even when a deductible has not been met and even when the person was eligible for Part A or Part B but declined to enroll.
- Part B's coverage regulation lists services, not drugs; most Part B drugs arrive through the category of services and supplies furnished incident to a physician's professional services.
- The incident-to regulation defines services and supplies to include drugs or biologicals "that are not usually self-administered," furnished in a noninstitutional setting to noninstitutional patients.
- The Part B limitations regulation states that Part B does not pay for any drug or biological usually self-administered by the patient, with named exceptions for certain diagnostic services, blood clotting factors and EPO.
- A definitional exclusion from Part D and a formulary denial under Part D are different events, and only the second one engages a Part D plan's formulary and exceptions machinery.
Answer first: Part D is defined by what Part A and Part B already cover
The regulation that defines a Part D drug does two things. It lists what can qualify — a drug available only by prescription and described in specified provisions of the Medicaid rebate statute, a biological product, insulin, medical supplies associated with the injection of insulin, a licensed vaccine and its administration, supplies directly associated with delivering insulin into the body, and certain combination products regulated by FDA as a drug, vaccine or biologic. All of that is conditioned on the drug being used for a medically accepted indication.
Then it subtracts. A Part D drug does not include any drug for which payment "as so prescribed and dispensed or administered to an individual is available for that individual under Part A or Part B." The exclusion is written to bite even in the awkward cases: it applies "even though a deductible may apply, or even though the individual is eligible for coverage under Part A or Part B but has declined to enroll in Part A or Part B."
That is the whole boundary. Part D does not begin where Part B's generosity runs out; it begins where Part B's authority to pay ends. If Part B could pay for the drug as it was prescribed and administered, the drug is outside Part D by definition, and a Part D plan is not the payer — whether or not the person actually enrolled in Part B.
What Part B pays for is a list of services, not a list of drugs
Part B's coverage regulation does not open with a drug formulary. It opens with categories of "medical and other health services": physicians' services; services and supplies furnished incident to a physician's professional services, of kinds that are commonly furnished in physicians' offices and are commonly either furnished without charge or included in the physicians' bills; diagnostic laboratory and X-ray tests; medical supplies, appliances, and devices; durable medical equipment; ambulance services; and a long list of other items, including specified vaccines and their administration.
Drugs reach Part B mostly through that second category. The "incident to" regulation defines the phrase "services and supplies" to mean any services or supplies "including drugs or biologicals that are not usually self-administered" that fall within the relevant statutory benefit and are not separately listed as their own Medicare benefit. It then states that Medicare Part B pays for services and supplies incident to the service of a physician or other practitioner, and it requires that they be furnished in a noninstitutional setting to noninstitutional patients, along with supervision conditions the rule spells out in detail.
So the usual Part B drug is one administered as part of somebody's professional service, rather than handed to a patient to use later. That framing, not a list of molecules, is what makes a drug a Part B item.
The self-administration limit is stated as an exclusion
The complementary rule appears in the section titled "Limitations on drugs and biologicals." It opens flatly: Medicare Part B does not pay for the following. First on the list, subject to named exceptions for certain outpatient diagnostic services, for blood clotting factors, and for EPO, is "any drug or biological which is usually self-administered by the patient."
Read the qualifier carefully. The word is "usually." The test is not whether one particular person self-administered one particular dose; it is a general characterization of how the drug is ordinarily taken. That is why the boundary is stable across patients rather than negotiated case by case.
The same section also excludes a narrow historical category — certain pre-1962 prescription drug products that FDA proposed to withdraw as less than effective for all their labeled indications and that are not presently subject to an FDA determination of compelling medical need — along with products that are identical, related or similar to those. That second exclusion is a legacy of the efficacy review program and rarely matters to a current retail prescription, but it is part of what the regulation says Part B will not pay for.
Why the exclusion reaches people who never enrolled in Part B
The most counterintuitive clause in the Part D definition is the one about declining to enroll. It means that whether a drug is a Part D drug is judged by what Part B could pay for, not by what the individual arranged for themselves.
The practical consequence is that a person who skipped Part B does not thereby convert a Part B drug into a Part D drug. The drug remains outside the Part D definition, and the Part D plan remains not the payer for it. The regulation reaches the same result for a deductible: the fact that a Part B deductible has not been met does not make Part B payment "unavailable" in the sense the definition uses.
This is a coverage rule with a design purpose. It keeps the two benefits from overlapping, and it keeps a beneficiary's own enrollment choices from moving costs between them. It also means that a rejection at a pharmacy counter is sometimes not a formulary decision at all but a statement that the claim was presented to the wrong benefit.
The definition contains two further exclusions that are separate from the Part A and Part B one, and they are worth knowing because they are sometimes mistaken for it. A Part D drug does not include drugs or classes of drugs, or their medical uses, that may be excluded from coverage or otherwise restricted under Medicaid by the provisions the regulation cites — with an express exception for smoking cessation agents. Nor does it include medical foods, which the regulation defines as a food formulated to be consumed or administered enterally under the supervision of a physician, intended for the specific dietary management of a disease or condition for which distinctive nutritional requirements are established by medical evaluation, and not regulated as drugs under the cited provision of the Federal Food, Drug, and Cosmetic Act.
Those three exclusions do different work. The Part A and Part B one routes a drug to another benefit. The Medicaid-derived one and the medical foods one route it nowhere: a product excluded on either of those grounds is outside Part D without being inside Part B, which is a materially different answer for anyone trying to work out who pays.
Where the boundary shows up in practice, and how to read it
A Part D plan's obligations begin only once a drug is inside the Part D definition. The access regulation for covered Part D drugs runs to formularies, tiering, transition supplies, exceptions and the machinery a plan must operate for drugs it covers. None of that apparatus is engaged by a drug the definition has already placed outside Part D.
That is why two different-sounding answers can both be accurate about the same medication. "Your plan does not cover it" is a formulary statement about a Part D drug. "This is not a Part D drug" is a definitional statement, and it points toward Part B rather than toward a formulary exception request.
The questions that separate the two are narrow. Is the drug ordinarily administered by a clinician as part of a professional service, or ordinarily taken by the patient? Was it furnished incident to a physician's service in a noninstitutional setting? Is it one of the items Part B lists as its own benefit, such as a specified vaccine? A drug that answers yes to the administration questions is being examined under Part B's rules; a drug that is usually self-administered is excluded from Part B payment by the limitation regulation, which is what leaves room for it inside Part D.
None of these determinations belongs to a marketing page. They are made by the payer applying the regulations to the specific product, the specific setting and the specific way the drug was prescribed and dispensed or administered — which is the exact phrasing the Part D definition uses.
Sources
- 42 CFR § 423.100 — Definitions (Medicare Prescription Drug Benefit; definition of "Part D drug")The categories that qualify as a Part D drug and the requirement that they be used for a medically accepted indication; the exclusion of drugs for which payment is available under Part A or Part B, including the clause covering a deductible and a declined enrollment; and the exclusions for drugs restricted under Medicaid, with an exception for smoking cessation agents, and for medical foods.
- 42 CFR § 410.10 — Medical and other health services: Included servicesThat Part B coverage is framed as a list of medical and other health services, including physicians' services; services and supplies furnished incident to a physician's professional services of kinds commonly furnished in physicians' offices; diagnostic tests; medical supplies, appliances and devices; durable medical equipment; and specified vaccines and their administration.
- 42 CFR § 410.26 — Services and supplies incident to a physician's professional services: ConditionsThe definition of "services and supplies" as including drugs or biologicals that are not usually self-administered, and the requirement that Part B incident-to services and supplies be furnished in a noninstitutional setting to noninstitutional patients under the supervision conditions the section states.
- 42 CFR § 410.29 — Limitations on drugs and biologicalsThat Medicare Part B does not pay for any drug or biological usually self-administered by the patient, subject to the exceptions stated for certain outpatient diagnostic services, blood clotting factors and EPO, and the further exclusion of certain pre-1962 products subject to an FDA notice of opportunity for hearing and of products identical, related or similar to them.
- 42 CFR § 423.120 — Access to covered Part D drugsThat the formulary, tiering, transition and access requirements a Part D sponsor must operate apply to covered Part D drugs, which is what distinguishes a formulary decision from a definitional exclusion.
Frequently asked questions
What makes something a Part D drug?
The regulation lists the qualifying categories — a prescription-only drug or a biological product described in specified provisions of the Medicaid rebate statute, insulin, medical supplies associated with injecting insulin, a licensed vaccine and its administration, supplies directly associated with delivering insulin into the body, and certain FDA-regulated combination products — each used for a medically accepted indication. It then excludes drugs for which payment is available under Part A or Part B, drugs or classes that may be excluded or restricted under Medicaid with an exception for smoking cessation agents, and medical foods as the regulation defines them.
Does declining Part B move a drug into Part D?
No. The definition excludes a drug from Part D if payment for it is available under Part A or Part B "even though a deductible may apply, or even though the individual is eligible for coverage under Part A or Part B but has declined to enroll." The test is what Part B could pay for, not what the person enrolled in.
What does Part B say about drugs a patient takes themselves?
The Part B limitations regulation states that Medicare Part B does not pay for any drug or biological which is usually self-administered by the patient, subject to exceptions it names for certain outpatient diagnostic services, for blood clotting factors, and for EPO. The operative word is "usually," which makes the test a general characterization of how the drug is ordinarily taken rather than a fact about a single dose.
How do drugs get paid under Part B at all?
Mainly through the "incident to" category. Part B's list of included services covers services and supplies furnished incident to a physician's professional services, of kinds commonly furnished in physicians' offices. The incident-to regulation defines services and supplies to include drugs or biologicals that are not usually self-administered, and requires that they be furnished in a noninstitutional setting to noninstitutional patients under the supervision conditions the rule sets out. Certain vaccines and other named items appear on Part B's included-services list in their own right.
Is "not a Part D drug" the same as "not on the formulary"?
No, and the difference changes what can be done about it. A formulary decision is a plan decision about a drug that is inside the Part D definition, and the Part D access rules — formularies, tiering, transition supply, exceptions — apply to it. A definitional exclusion says the drug is outside Part D entirely, so the Part D machinery is not what governs it.
Who decides which part applies?
The payer, applying the regulations to the specific product and the specific circumstances. The Part D definition turns on the drug "as so prescribed and dispensed or administered to an individual," which means the setting and the manner of administration are part of the determination. This article describes the rule; it does not determine any product's classification, and no classification for a specific medication was researched for it.