Why the eligibility check decides whether the episode is payable
A home health claim rarely denies because the care was unnecessary. It denies because the agency could not show, on the start-of-care date, that the patient met Medicare's conditions of eligibility — or because the payer on the claim was not the payer that actually held the risk that day. Both failures are found by a check that costs minutes and prevents a loss that costs months.
Timing is what makes it matter. Verification happens before a clinician drives to the house, before the OASIS assessment is completed and locked, before a 30-day period is priced. Everything afterwards is either correctable at a cost in clinician time or not correctable at all: a benefit period that ran out on the start-of-care date cannot be restored with better documentation, and a visit delivered under an authorization that never existed is a write-off with a compliance finding attached.
What follows treats verification as a checklist rather than a formality: what Medicare actually requires, what changes when the payer is a Medicare Advantage plan, which fields of the 270/271 response decide the outcome, and what to record so the answer still exists when a denial arrives six months later.
The four conditions Medicare actually requires
Medicare does not pay for home health because a patient is elderly, chronically ill, or carries a diagnosis that sounds like it belongs in home care. Payment depends on four conditions, and each one has to be evidenced in the record at the start of care — being true in the clinician's judgment is not the same as being supportable from the chart.
The conditions are cumulative. An agency can be comfortable that a patient is homebound, and still lose the period because the plan of care was never signed, or because the face-to-face encounter was dated outside its window. Verification is the point where all four get checked at once, before any of them becomes expensive.
Condition one: the patient is confined to home
Homebound does not mean bedbound. It means leaving home requires considerable and taxing effort, or is medically contraindicated. Absences are permitted when they are infrequent, of relatively short duration, or to receive medical treatment, so a standing outpatient schedule does not disqualify a patient.
What kills the claim is a record that cannot describe why leaving home takes considerable effort: the criterion asserted as a conclusion (patient is homebound) rather than supported (patient becomes short of breath walking to the mailbox, uses a walker, has not left the apartment in three weeks except for a physician visit). The verification question is whether the note on file already answers it.
Condition two: the need for skilled services is intermittent
The plan has to call for skilled nursing, physical therapy or speech-language pathology on an intermittent basis; occupational therapy counts toward eligibility only while the plan still requires it. Intermittent means the skilled need is not full-time and is expected to end, or to be needed only periodically.
A patient whose remaining need is custodial — bathing, dressing, meals, supervision — is not eligible however frail they are, and breadth of documentation does not change that arithmetic. When the referral reads as a custodial need, the honest move at intake is a conversation with the referral source, not a plan of care written to fit.
Condition three: a signed plan of care reviewed by a practitioner
Orders must be signed and dated, describe the services with their frequency and duration, and be reviewed at least every 60 days through recertification. This is the condition most often broken by process rather than judgment: the care is right, the orders are verbal and unsigned, and the episode is technically uncertified.
An order that arrives after the claim does not repair the certification window it missed. Verification should be able to name the order on file and its date before the start of care is scheduled, which is also the moment when chasing a signature is still cheap.
Condition four: a face-to-face encounter supports the eligibility
The certifying practitioner, or an allowed non-physician practitioner, must document an encounter related to the primary reason the patient needs home health, within the 90 days before the start of care or the 30 days after it.
At verification the question is narrow: does the encounter exist, is it dated, and is it about the reason home health is starting — not an unrelated office visit that happens to fall inside the window. A missing encounter is an agency problem even though the agency did not create the note, because the agency is the one that has to produce it.
What changes when the payer is a Medicare Advantage plan
For traditional Medicare the check is mostly about the patient: coverage active, benefit available, conditions met. For a Medicare Advantage plan the check is about the patient and the contract, and four separate questions have to be answered before the start of care.
Plan effective date. The member is enrolled, but the plan's coverage may not begin until a date after the referral. A start of care that precedes the effective date is billable to nobody.
Network status. An out-of-network agency may be paid at a different rate, or not at all, depending on the plan type and benefit.
Prior authorization. Traditional Medicare fee-for-service does not require prior authorization for home health; Medicare Advantage plans can and routinely do. Verifying eligibility without verifying authorization produces a clean claim that denies for a reason the eligibility response never mentioned, which is the single most common source of avoidable Medicare Advantage write-offs.
Authorized scope. An authorization for skilled nursing visits is not an authorization for physical therapy, and one for six visits is not one for twelve. Units, disciplines and the covered period have to match the plan of care before the first visit is made.
That last point is where most Medicare Advantage revenue leaks: the authorization exists, the visits exceed it, and the excess is discovered at the denial instead of at the schedule. Medicare Advantage home health billing covers the authorization and appeal mechanics in more depth.
Reading a 270/271 response instead of glancing at it
The eligibility inquiry and response — the 270 and 271 transactions under HIPAA — return coverage status and benefit detail. The failure mode is reading the first line, seeing coverage active, and moving on. Five items in the response change what you can bill.
Coverage on the start-of-care date, not on the date of the inquiry. A plan that terminated last month and a plan that starts next month both answer an inquiry with a member record.
Part A versus Part B. Home health is covered under Part A, or under Part B for a patient who is not entitled to Part A. Which one the patient holds changes the benefit being drawn on.
The home health benefit itself, including how much of the current period remains. A patient can be eligible, enrolled and out of home health benefit at the same time.
A hospice election. It changes what may be billed under Medicare and belongs in the verification record rather than in a billing surprise three months later.
Another payer being primary. Where a group health plan, workers' compensation or another insurer is primary, the Medicare claim is secondary and the verification has to name the primary payer.
The 271 is a data source, not a decision. It answers whether the patient has coverage; it does not answer whether the patient is homebound, needs skilled services, or has a signed plan of care. Agencies conflate the two and then treat an active eligibility response as clearance to start care.
Where verification actually breaks
The same five failures account for most eligibility-driven denials, and none of them is a payer problem.
Checking once, at referral. Plans change, effective dates move and authorizations expire between the referral and the start of care. The check that counts is the one dated the day the start of care is scheduled, or the day before.
Verifying eligibility and assuming authorization. Two separate responses from two separate systems, with two separate ways to be wrong.
A verbal confirmation with no record. Someone at the plan said it was fine. Six months later nobody can produce who said it, when, or exactly what they confirmed.
Trusting the stored insurance. A record that still says Medicare Advantage for a member who moved to traditional Medicare — or the reverse — produces a claim to the wrong payer with a correct-looking patient file behind it.
Scheduling the start of care before the benefit is confirmed. The visit happens, the OASIS is completed, and the question of who pays becomes a conversation instead of a check.
What to write down, and why it matters in an appeal
A verification record that survives an appeal contains the date and time of the check, the payer name and ID, the plan type, the coverage effective date confirmed, the benefit status returned, the authorization number with its units, disciplines and date range, the portal or transaction reference, and the name of the staff member who performed it.
That record answers the question a denial will eventually ask: was this episode verified against the payer that held the risk, and was the service authorized? An eligibility denial is one of the few that is genuinely preventable, because the answer existed before the first visit. When it was documented, the appeal is an attachment; when it was not, the appeal is a reconstruction, and reconstructions lose.
It also matters on the receivables side. Denials that trace to a verification gap are process defects, not payer disputes, and they should not be worked in the same queue as clinical denials, because the fix is upstream. Home health AR follow-up depends on that separation: a team that appeals a preventable denial without repairing the check will appeal the same denial again next month.
Where the check belongs in the intake sequence
Sequence matters more than thoroughness, because each step depends on the one before it.
Referral. Payer named by the referral source and verified independently rather than trusted on the strength of the referral.
Eligibility. Checked against the intended start-of-care date, on the payer that will actually be billed.
Authorization. Requested where the payer requires it, for the disciplines and visit counts the plan of care will use.
Start of care. Scheduled only after the two steps above return an answer.
Assessment and notice. OASIS completed and reviewed, and the Notice of Admission filed inside its timely-filing window.
Claim. Submitted and priced on the assessment as documented.
The Notice of Admission is the step most often lost between intake and billing, and it is mechanical rather than clinical. The NOA timely-filing rules are worth pinning where the intake team can see them, because a late notice reduces payment on a period that was otherwise clean.
None of this makes an episode eligible that was not. Verification makes the case that it was eligible answerable on the day someone asks, with records instead of recollection. At volume the whole chain — verification, authorization, coding, notice of admission, claims, denial and AR follow-up — is what an outsourced revenue cycle exists to run end to end, and verification is simply the first place where it either works or does not.