Patient assistance program eligibility, explained.

A free medicine application, followed from the doctor's office to the pharmacy

1. The doctor prescribes and coverage falls shortAna's plan upholds its denial on appeal on September 24, 2026.
2. The office and the patient send a packetAn 11-page fax with six documents reaches the program on October 6 at 2:12 pm.
3. A reviewer checks every rule against the pagesResidency, prescription, coverage path, household, income, signatures, current proof. The household question is settled here, or settled wrongly.
4. A person decides, and a second person signs where the rules askAna's approval needs one signature. In this program, a denial or an approval under Medicare Part D, the drug benefit, needs two.
5. The letter, the pharmacy and next year's renewal followFree medicine ships, and a reminder is set before the enrollment ends.

Step 3 is the subject of this guide. The people, program and figures in this example are fictional.

Four surprises in a poverty guideline eligibility review

Four terms to know before reading a free drug application

Free drug programs, copay cards and charity funds: who each one serves

In shortA manufacturer's free drug program gives product, usually to people with no usable coverage and limited income. Copay cards and charity funds help insured people pay their share, and the federal rules treat each one differently.

People use "patient assistance" for several programs that work in different ways. A review in JAMA sorted the ways to afford a prescription: copay cards for privately insured patients on brand-name drugs, patient assistance or public programs for uninsured people or those on public insurance such as Part D who meet financial criteria, plus coupons and direct purchase.6 CMS, the federal Medicare agency, describes manufacturer programs as giving "financial assistance or drug free product (through in-kind product donations) to low income individuals".7 This guide is about the free product decision.

That explains why the free drug decision carries so much weight for someone like Ana. If she had no insurance at all, most of the charity programs in that study would not have taken her.8 With a plan that has finally refused the medicine, a copay card has no share to reduce. One caution on the study: it is from 2018 and covers the charities, so read it as a picture of how those programs were built. We found no newer count.

HHS-OIG bulletins and advisory opinions on patient assistance, in date order

In shortNo federal statute sets out how to run a free drug program. The rules come from the anti-kickback statute as read in two OIG bulletins, a set of advisory opinions that bind only the company that asked, CMS's Part D guidance and the yearly HHS guideline notice.

The anti-kickback statute makes it a crime to knowingly and willfully offer or pay anything of value to induce business paid for by a federal health care program.3 Giving a Medicare patient a free drug could look like that, so OIG, the HHS inspector general's office, has explained in bulletins when the risk is low.

A company can also ask OIG for an advisory opinion on its own facts. An opinion rests on what the company told OIG, protects only that company, and "may not be relied on by any persons other than" the requestor.3 Read opinions to learn what OIG looked at, and take your own program to your own counsel.

OIG's compliance page files its 2003 guidance for drug makers under an archive. It lists a general compliance guide from November 6, 2023 as current, and gives no date for a new guide for drug makers.23 A reviewer's checklist therefore rests on the 2005 bulletin and the opinions above, and our search of OIG's index on November 3, 2026 found no newer bulletin on manufacturer free drug programs.24

The 2026 HHS poverty guidelines, and what 400% of them means in dollars

In shortFor 2026 the figure is $15,960 for one person plus $5,680 for each added person, in the 48 contiguous states and DC. Alaska and Hawaii have their own tables. A 400% limit for a family of four is $132,000.

HHS published the 2026 guidelines in the Federal Register on January 15, 2026, effective January 13 "unless an office administering a program using the guidelines specifies a different effective date".1 They reflect a 2.63% price rise between 2024 and 2025.1 HHS keeps two measures: the Census Bureau's poverty thresholds, for statistics, and these guidelines, a simplified version for deciding eligibility.25 ASPE, the HHS office that publishes them, names pharmaceutical companies among the private users.25

You will hear "federal poverty level" or "FPL" everywhere. HHS calls that phrase ambiguous and advises against it, particularly in legislation or rules.26 Most program forms still use it, so this guide says "guideline" and accepts that your form may say FPL.

Guideline figures from HHS.26 The right-hand column is our multiplication, and 400% is one program's example limit. HHS sets no such limit.

The guidelines are not defined for Puerto Rico or other outlying jurisdictions, so a program that serves them has to say what it uses.1 A program serving all 50 states needs the Alaska and Hawaii tables in its rules, along with a date for switching to each new year's figures.

Switching the table in January can flip a decision

Take a fictional family of four earning $130,000. On the 2025 guideline of $32,150 they sit at 404%, over a 400% limit. On the 2026 guideline of $33,000 they sit at 394%, under it.

A program that decides that family's application during January needs a written answer to which table applies: the date the application was received, the date it was decided, or the date the program adopted the new notice. HHS lets a program set its own effective date.1 On our reading, any of the three works if it is applied the same way to everyone. Confirm your choice with counsel.

How to turn pay stubs, Schedule C and benefit letters into household income for a PAP

In shortAnnualize each source by its pay frequency, add every household source the rules count, divide by the guideline for the household size, and compare before rounding. Whether income means gross or net is your program's written choice.

HHS is plain about this: "There is no simple answer" to whether income is before or after tax, and programs define it in different ways.25 OIG's Advisory Opinion 25-01 describes a manufacturer that checks income by a soft credit check or by reviewing documents.4 That opinion binds only its requestor, and whichever method a program picks, OIG's test is an even measure of need.5 Our example uses documents and gross income, as the sample program's rules do.

Ana Ruiz's household income, step by step

Now add the $20,120 back. Read as weekly, Ana's household is at 657% and she would be wrongly denied. Read as twice a month it is 336%, still under, with the wrong figure on file for an auditor. Pay stubs usually state the period, but a faxed copy may blur it, and the two dates on consecutive stubs are the check: 14 days apart means every two weeks.

Five choices your written rules have to make about income

Who counts in the household when the PAP application and the tax return disagree

In shortThe return looks back at last year and the application describes today. A birth, a death, a divorce or a parent moving in can make them differ, and a written rule, applied the same way and recorded each time, settles it.

The tax return is good evidence of income for a past year. It is weaker evidence of who lives in a home this October. Ana's 2025 return names three people because Leo was born in March 2026. Her application, which she attested to (signed as true), names four. Income does not change, but the guideline does, and the result crosses the limit.

A reviewer who treats any mismatch as a reason to request more paper delays a family with an obvious answer, and a reviewer who always trusts the application will one day accept a count nobody can support.

A useful rule says three things: whom it counts (for example, the people living in the home today, as attested), what evidence would make the reviewer ask (a return showing fewer people with no birth or move to explain it), and what the reviewer must write down when the two differ. HHS leaves the definition of family or household to your program, so the record of your reasoning is your defense.1

Uninsured, denied after appeal, Medicare or Medicaid: the coverage path each applicant takes

In shortBefore income, a program asks whether something else should pay. Each answer calls for different proof, and a denial at prior authorization, the plan's approval step before it will pay, often falls short of what a program asks for.

In OIG's Advisory Opinion 06-03, the manufacturer's program first helps an applicant explore other help, such as private insurance, veterans' benefits, Medicare or Medicaid, and directs the person there when it is available and appropriate.3 Its Medicare applicants who had not enrolled in Part D had to apply for the low-income subsidy and be rejected before qualifying.3 Two other opinions asked people below 135% of the guideline to show they had been denied or were ineligible for that subsidy.2,18

Extra Help now reaches further: for 2026 the income limits are $23,940 for one person and $32,460 for a married couple, which match 150% of the guideline (our arithmetic), with resource limits of $18,090 and $36,100.10

This is the area most likely to change in 2027. CMS's interim final rule requires 80 hours a month of work or similar activity for many adults in the Medicaid expansion group, in 43 states and DC, with states generally starting no later than January 1, 2027.14 People who lose coverage may then apply for free medicine with a packet that describes a coverage history a program's forms were not written for.

Medicare Part D and patient assistance: six OIG conditions and a spending test the cap can overtake

In shortFree medicine for a Part D enrollee must sit entirely outside the benefit. OIG listed six conditions in 2005. The out-of-pocket threshold ($2,100 in 2026, $2,400 in 2027) means a 3% spending test written in 2006 can fail some higher-income households on Part D costs alone, by our arithmetic.

Part D has a yearly limit on what an enrollee pays: the out-of-pocket threshold, often called the cap. TrOOP, or true out-of-pocket spending, is what moves a person toward it. OIG's fear in 2005 was a drug maker paying a person's share to speed them toward the point where Medicare pays. So it treats free drugs differently from subsidized copays: "no claims for payment" may go to the plan or the patient, and the help "must not count toward the beneficiary's TrOOP or total Part D spending for any purpose".5

The six conditions OIG named for free drugs outside the Part D benefit

1. Plans are told. Safeguards ensure Part D plans know, so no plan pays and nothing counts toward TrOOP.5
2. The whole coverage year. Help lasts for the Part D coverage year, or what remains of it after the first fill.5
3. Periodic use is covered. Help stays available even if the person uses the drug only now and then.5
4. Records are kept. "Accurate and contemporaneous records" let the government verify that the drugs were outside the benefit.5
5. Need is measured evenly. "Reasonable, uniform, and consistent measures of financial need", without regard to "the providers, practitioners, or suppliers used by the patient or the Part D plan".5
6. CMS guidance is followed. CMS's 2006 memorandum allows a nominal copay, says a claim is never submitted to the plan, and the rule on incurred costs lets a documented nominal copay count.19,21 CMS's data sharing agreement has the agency flag the program as a payer that does not count toward TrOOP.20

OIG said these points reduce risk. It also said the application of the law is difficult to assess "absent a specific set of facts".5 Six conditions are a starting list for your counsel, and meeting them gives no protection on its own.

Why a 3% spending test needs a second look after the cap

In Advisory Opinion 06-03, Part D enrollees had to have spent 3% of household income on outpatient prescription drugs that year, on top of an income limit of 325%.3 Advisory Opinion 07-04 describes a stiffer one: 4% incurred and 10% anticipated.18 The idea was sound when a Part D enrollee could keep paying well into a year. Since 2025, the enrollee's costs on covered drugs stop at the threshold: $2,000 in 2025, $2,100 in 2026 and $2,400 in 2027.13,12 So the test and the cap can now collide.

The sums are ours, and no agency has published them: 3% of $70,000 is $2,100, and 3% of $80,000 is $2,400.12 How much this matters depends on the limit. At Advisory Opinion 06-03's 325%, a couple tops out at $70,330 on the 2026 guideline, so the collision barely appears. A single person at a 400% limit tops out at $63,840, where 3% is $1,915, under the threshold. A married couple at a 400% limit earns up to $86,560, and 3% of that is $2,597, more than either threshold.

OIG does not require the test: in Advisory Opinion 07-04 it said it would likely reach the same outcome if that requestor dropped its spending test.18

Spending on drugs outside the benefit can still count if your rules say so, so read this as a question for your rules: what does the test measure now, and does it screen out people it was written to help? A person's yearly cost also spreads out if they use the Medicare Prescription Payment Plan,28 so "spent this year" and "incurred this year" may differ on the statements a reviewer receives. Check that wording too.

Two more points belong in your rules. A 12-month enrollment that runs across two coverage years sits awkwardly with an OIG condition framed around the coverage year, and with the opinions in which eligibility is reassessed each year and federal enrollees re-apply each calendar year.3,4 And anyone near the Extra Help limits should be asked about it before a free drug decision, as the opinions show.3,2,18

Ten eligibility errors that deny the wrong patient or approve the wrong one

In shortErrors cluster around pay frequency, household counts, the wrong coverage letter and out-of-date proof. We found no published error rate, and an audit of your own past files is the way to get one.

Research on these programs looks at the applicant's side. A 2025 scan of 66 affordability resources for five conditions found varied criteria, varied application steps and requirements for "sensitive financial documents".29 A 2024 review of oral cancer drug programs names "limited transparency on eligibility criteria" and "onerous documentation burdens".30 A 2025 review of five articles called the findings "mixed" on what would qualify a patient.31 None of them counts reviewer errors. The list below is ours, drawn from the steps above: an analysis, for you to test against your files.

What an uneven PAP decision risks: patients without therapy, kickback exposure, best price and privacy

In shortThe first cost falls on the patient who waits or is wrongly refused. For the program, the risks are the anti-kickback statute, Medicaid best price, and the handling of tax returns and health information. The large public settlements so far concern copay charities.

The person left waiting

The patient carries this risk. A 2025 review names prescription-to-start delays among the barriers these programs create.31 A person whose plan has already said no, twice, has run out of other doors.

The anti-kickback statute and the settlements people quote

Press stories about "patient assistance" settlements usually concern something other than a free drug eligibility review. The three announcements below are about drug makers and copay charities, and each is described on OIG's enforcement page as resolving claims or allegations that companies used foundations as conduits to pay Medicare copays.32,33,34 We found no public case that turned on a reviewer's free goods eligibility decision.

They still teach a lesson about free goods. OIG's bulletins ask whether help is steered by something other than need. A free drug program's equivalent is a decision that takes account of who prescribed, which pharmacy would dispense, or which plan the patient chose. Opinions 06-03, 06-14 and 25-01 each record that the requestor said its decisions were made without regard to such things.3,2,4 A program should be able to show the same.

Best price, HIPAA and credit checks: three more laws a free drug program meets

A packet holds a tax return, pay stubs, a diagnosis and a Social Security number. The practical rules are simple: mask tax and account numbers when reading, keep line-level income only as long as your rules require, and keep income line items out of letters.

How big manufacturer patient assistance foundations are, and who may need them next

In shortSeven drug makers' foundations we looked up reported about $19.5 billion in total expenses for tax year 2023, by our sum. They are a selection, so that is no market total. Uninsured people numbered 26.7 million in 2025, and federal estimates point up.

These are filings by the foundations we looked up. Other drug makers' foundations file too, so the chart is no ranking and its sum is no market total.

Year-to-year swings can be large: one foundation in the chart reported $1.80 billion for 2022 and $2.97 billion for 2023.38 We read the data totals, not the filed forms, so we cannot say how much of each total is donated medicine, how the filer valued it, or how many patients it reached. We found no published total for free goods, for applications a year, or for approval rates across the industry.

What the field does publish is small and local. A West Virginia health system's in-house team assessed 153 patients and got no-cost medicine for 91 of them: 283 fills worth $2,060,633.83 in total.41 A systematic review of 33 studies in 2011 found limited information on effectiveness and called for more rigorous research.42

Why demand is likely to change in 2027

Three shifts are in view. Medicaid work rules begin in most states by January 1, 2027.14 KFF's allocation of CBO's estimates puts about 10 million more people uninsured in 2034 from the 2025 law, and about 14.2 million counting the enhanced tax credits' expiry.40 We read KFF's brief, not CBO's own tables.

Meanwhile the Part D threshold rises to $2,400 in 2027, and OIG has shown it wants to revisit charity arrangements as the cap takes hold.12,13 We make no forecast of applications. Your own monthly counts, split by coverage path, are the figure to watch.

What software can do in a PAP eligibility review, and what state AI laws ask about denials

In shortSoftware can sort pages, read figures, do the arithmetic and compare decisions across files. A person should decide every denial. California and Colorado rules on automated decisions start January 1, 2027. Whether they reach a free medicine denial is an open question.

State rules on automated decisions that may reach eligibility denials

Two states have written rules that may reach a denial. They differ on what sets them off and on who stands outside them.

The Colorado attorney general's rules are due by January 1, 2027.16 A law firm's note of October 23, 2026 says they were not yet final, and that under an April 27 court order the attorney general will not enforce against violations occurring up to 14 days after a ruling on a motion still to be filed. We read that note, not the order.44

Neither text says whether a free medicine denial counts as healthcare services, and we found no ruling on it. That is an open question for counsel, and this guide takes no position. A sound design answers it anyway: a named person reviews every denial, can see why, and can change it, and the applicant gets a route to ask for another look.

Why FDA's AI draft and drug quality rules sit outside this decision

FDA's January 2025 draft guidance on AI concerns regulatory decision-making for drugs and biological products, and FDA's page still shows it as a draft on November 3, 2026.45 Choosing who gets free product is a commercial eligibility decision, so on our reading that draft, and quality-system texts for drug records and manufacturing such as 21 CFR Part 11, draft EU GMP Annex 22 and GAMP, are not the texts that govern it.

Your validation practice may still borrow from them. No software makes a program lawful by itself: controls can support your compliance, and the program remains yours to answer for.

How free goods teams run eligibility today: from the fax line to the pharmacy referral

In shortAn intake clerk logs and sorts, a reviewer types in the figures (keying) and decides, a senior reviewer signs the harder ones, a pharmacy ships. In the sample program's estimate, most of the time goes into keying and checking. Some programs outsource all or part of it to a hub vendor, a company that runs patient support for drug makers.

The roles around a free medicine application

Three ways to staff a free goods review

Ana's household question and a fax that lands at 8:42 am, from packet to signed approval letter

In shortSam Patel, an eligibility reviewer, opens the day with 20 open applications. One fax is read while he watches. One application waits on the only question a person must answer. He answers it, signs, and the letter and the pharmacy referral follow, with the record kept.

1

The application's route through the rules, already lit

The rail on the left sorts the day's applications by what they need. Ana Ruiz's application for Veltrimab, PAP-10418, sits under "Needs you", showing both results: 359% or 433%. In the center is the program's rule list drawn as a path: lives in the U.S., approved use and U.S. prescriber, coverage path, who lives in the home, income within 400% of the guideline, signed and current.

Each passed step names the document and page behind it, such as "Prescription p. 5". The insurer's denial, upheld on appeal, sits on the "Denied after appeal" branch. The two amber steps both turn on one question.

2

A new fax is read while the reviewer watches

Marisa Santillo's packet for Talvoren arrived at 8:42 am: 8 pages, 4 documents, from Dr. Tomás Ibarrondo's office in Houston. Sam clicks "Read the packet". The Packet Reader labels the pages and the path lights in order as each agent finishes its check, with the step under way named above the path.

3

$1,472 a week, 280% of the guideline, and the pay stub beside it

Her pay stub from Ship Channel Terminals, dated October 2, shows $1,472.00 weekly. Multiplied by 52, that is $76,544, against $27,320 for a household of three: 280%, under the 400% limit. All 13 of the program's rules are met, and the recommendation reads "Eligible, recommend approve". The source page sits on the right with the deciding figure marked.

4

The application says four, the tax return says three

The full review shows the conflict side by side. The application, page 2, lists Ana, her husband Tadeo, Sofia, 9, and Leo, born in March 2026. The 2025 return shows three. Income is unchanged: $3,780 every two weeks × 26 = $98,280, plus $20,120 from Tadeo's Schedule C, for $118,400.

The program's rule 4.2 counts the people living in the home today, as attested and signed. It asks the reviewer to record which count was used and why. That rule belongs to the sample program, in keeping with HHS leaving such definitions to each program.1

5

Sam counts four, the path turns, and he signs

He counts four. The household step records "4 people · counted by Sam Patel", the income step passes at 359%, and the outcome becomes eligible. The approval shows Veltrimab free for 12 months, October 7, 2026 to October 6, 2027, shipping from Clearwater Free Goods Pharmacy. The basis is four people, 359%, 13 of 13 rules. This case needs one signature. The letter draft takes every figure from the decision.

He signs on October 7 at 9:13 am, 0.9 business days after the fax, on the screen's clock. By the program's 45-day setting, the reminder falls on August 22, 2027 (our date arithmetic).

Denials, Medicare Part D, Medicaid and missing paperwork on the same rule path

In shortThe same rules send other applications down other branches. A denial needs a second signer. A Part D approval needs a second signer and a spending check. A Medicaid card ends the path. A blank signature brings a request.

Over the limit: Bradley Hulme at 612%, and a second signer

Bradley Hulme applied for Dravenzo after Front Range Mutual upheld its denial. His 2025 return shows $167,200 for a household of three, or 612%. The path leaves the rules at the income step, in red. The recommendation is a denial, and the program requires Omar Haddad, the senior reviewer, to second-sign before any letter goes out. A person decides each denial here, and a second person checks it.

Medicare Part D: Walter Brookhart's spending check and two signatures

Walter Brookhart, 71, has Part D through Prairie Senior Rx. A Social Security letter at $1,840 a month and a pension at $810 a month come to $31,800, which is 147% of the $21,640 guideline for two. He has spent $1,960 on prescriptions this year, 6.2% of income, above the program's 3% setting.

The screen says the free medicine is given outside the Part D benefit and the plan is told, as OIG's conditions expect.5 Omar Haddad signs after Sam, because the program's rules say so. OIG does not require two signers: this is the sample program's own control.

A Medicaid card, a blank signature: the path stops or asks

The Applications list, re-enrollment and turnaround views

What the Packet Reader and five other agents do, and where a reviewer and a senior reviewer sign

In shortEach agent has one job and a line it stays behind. The reviewer signs every approval and every denial. Denials and Part D approvals take a second signer. Compliance approves each version of the rules.

Who signs a free medicine decision in the sample program

The prescriber, the office and prescription volume are never inputs. Every agent step, signature and letter is on the activity trail with who and when. These controls support your compliance. They do not make a program compliant, and no software can certify one.

The reviewers and leads in the screens

From a faxed PAP application to a cited decision record, a letter and a pharmacy referral

In shortIt takes the packets you already receive and your written rules. It gives back a recommendation with its pages, a record for audit, letters from your templates and a referral. It connects to the systems you already run, by category.

The platform is SOC 2 Type II audited, with information security managed to ISO 27001. It keeps role-based access, maker-checker sign-off and a full audit trail, and scales across medicines, programs and reviewers on one set of versioned rules.

What keying, waiting and uneven decisions cost a free drug program

In shortYour costs are reviewer hours, packets sent back, days a patient waits, and the effort of proving even treatment. No public benchmark gives these for free goods programs, so you need your own numbers, and a week of sampling will find most of them.

Six PAP numbers you can pull in a week

Costs of a free goods program that no spreadsheet shows

What PAP automation returns: reviewer hours, fewer packets sent back and a record on every decision

In shortMost of the measurable value is reviewer time. The rest is fewer returned packets, faster decisions and audit evidence. The calculator shows the first two. Replace every default with your own number.

Six places a free goods program gets its return

Your hours and dollars across a year of PAP applications

The calculator favors higher-volume programs with long reviews. A program with a few hundred applications a year and short, simple rules may be better off fixing its forms and templates first. Dollars are also the weakest reason in this field: for many programs the stronger case is evidence and patient days, which you should report beside the money.

How ready your free drug program is: written rules, packet volume, proof of even treatment and who decides

In shortEight scored questions and one for reflection. A low score usually means writing down your rules comes first, which helps with or without software.

What each PAP readiness score tells you to do first

Questions that test PAP eligibility software or a hub vendor on income, household conflicts and proof

In shortAsk us too. Good answers show the working on your own packets. Warning signs are a single accuracy figure, a claim of compliance, and any path where software alone sends a denial.

Pilot PAP automation on past packets, with planted errors and a blind re-review

In shortMeasure first. Test on past packets with known answers and planted errors. Run in parallel with your reviewers signing every decision. Then decide with three numbers, and commit to nothing before that.

1. Write the rules down and have compliance approve the version you test: household, income, proof dates, coverage paths, the guideline table and its switch date.
2. Time 30 files and re-review 50 blind. That gives you minutes, days and a real difference rate.
3. Gather 200 past packets across every path, with masked identifiers and approval to share them. Plant known errors in a copy of 30.
4. Run in parallel on live packets for six weeks, with your reviewers deciding as usual and comparing afterward. Include a January-table test.
5. Decide with three numbers: errors caught by type, false alarms per 100 clean files, and reviewer minutes to a signed decision.

Five measures that show whether a PAP pilot worked

Patient assistance program questions answered: income limits, Medicare, household size and denials

In shortThe questions people search, each answered on its own. Program terms vary, so check the specific program's rules.