Eight Agencies, One 9:40 A.M. Referral: Why the Hospice Shortlist Now Happens Before Your Intake Phone Rings

Ask your intake coordinator how many referrals she never saw

She can tell you exactly how many came in. Every agency can. What she cannot tell you is how many times your agency was considered and dropped before anyone touched a phone. That number used to be small enough to ignore. In the last year it stopped being small.

Two different buyers decide where a home health or hospice patient goes, and they run on different clocks. The first is the hospital case manager or discharge planner, working a floor list, trying to clear beds before rounds. She sends one referral to six or eight agencies at once through WellSky CarePort or Aidin or Ensocare, usually between 9 and 10:30 in the morning, and the agency that responds with a firm accept and a start-of-care time wins. The second buyer is a daughter in another state with a hospital social worker's printout of four agency names and a laptop.

The daughter's behaviour is what changed. She no longer calls four agencies during her lunch break. She asks an assistant to do it, and the assistant reads whatever it can find about each agency, compares, and comes back with two or three. Whatever is not readable is not compared.

The eleven facts a shortlist actually turns on

Sit with a discharge planner for an hour and you learn that the decision is boringly factual. Before anyone asks whether you are compassionate, they need to know: the exact counties and ZIP codes you staff, not the ones on your license; which payers you take, including Medicare, the state Medicaid program, the specific Medicare Advantage plans, and VA Community Care; which levels of care you actually provide, which for a hospice means routine home care, continuous home care, general inpatient and respite, and where your contracted inpatient beds are; whether you take IV therapy, wound vacs, trachs, vents, TPN or peritoneal dialysis; whether you take pediatrics; which disciplines you can staff in that county this week, meaning physical therapy and occupational therapy as well as nursing; what languages your field staff speak; whether you accept an admission after 5 p.m. or on a Sunday; your typical hours from referral to start of care; your Care Compare star rating and CAHPS results; and whether you have an aide available, which in 2026 is the most common reason a referral falls through.

Now go look at your website. Most agency sites in this trade have a page that says "serving the community since 1998 with compassionate, patient-centered care," a service-area map picture with no text in it, a phone number, and a contact form that emails an inbox nobody watches on weekends. Every one of those eleven facts is either missing or trapped inside an image.

What changed: the shortlist is now assembled by software

Being findable in 2026 means your service area, payers, levels of care and response times exist somewhere a piece of software can read them without a human squinting at a picture — because the party researching you is increasingly another company's assistant, not a person with a browser tab.

Agent-to-agent protocols and machine-readable business surfaces landed properly this year. The practical effect for a home health or hospice owner is small and specific: an assistant working on behalf of a family, a hospital, or a Medicare Advantage plan's care management team can now read a business, compare it to others, and in some cases start the contact itself. That assistant does not scroll. It does not infer that "the greater metro area" includes the county across the river. It reads what is written plainly, and it skips what is not.

flowchart TD
  A["Case manager sends one referral to 8 agencies at 9:40 a.m."] --> D["Shortlist assembled"]
  B["Daughter asks her assistant to find hospices in the county"] --> D
  C["MA plan care manager screens in-network agencies"] --> D
  D --> E{"Are your ZIPs, payers and levels of care readable?"}
  E -->|No| F["Agency never appears in the comparison"]
  E -->|Yes| G["Agency appears with response time and star rating"]
  G --> H["Contact attempt: phone, web chat or referral portal"]
  H --> I["Intake confirms staffing and books the SOC visit"]

Tuesday, 9:40 a.m., with the facts written down

The referral hits CarePort: 81-year-old, CHF, discharging tomorrow to a home in a ZIP on the far edge of your territory, needs skilled nursing three times a week for two weeks, has a Medicare Advantage plan you contract with, daughter speaks Tagalog at home. Six agencies get it.

On your side, the intake coordinator's screen already shows the answer to the only question that matters: do you have an RN with capacity in that ZIP on Thursday. She accepts at 9:52 with a named start-of-care window. Two of the six never respond before noon because their intake person is on a school run.

That evening the daughter, three states away, asks an assistant which of the accepted agencies is worth choosing. The assistant reads your published service-area list and finds her mother's ZIP by number rather than by hope. It finds that you list Tagalog among your field-staff languages. It finds your star rating and your stated referral-to-start-of-care window. It puts you first on a list of three and tells her to call. Your line answers, because it answers at 8 p.m.

None of that required new clinical capability. It required writing down things you already know and were keeping in your head, in your scheduler's head, or in a PDF.

What one extra admission a week is worth

Assumptions, stated and illustrative: a home health agency admitting 60 patients a month; average payment per 30-day period around $2,050; an average of 1.7 periods per patient before discharge. Suppose being legible to the shortlist wins you four additional admissions a month — one a week — from referrals you were already receiving but losing on speed or on invisibility.

Run the same arithmetic on the hospice side using your own average length of stay and the routine home care rate. The shape does not change: the marginal admission is worth thousands, and the thing that lost it was usually a page nobody had updated since 2019 or a phone that rang out at 6:10 p.m.

What this will not do for you

It will not fix a two-star rating. If anything it makes a weak rating more visible, faster, because that is one of the first things a comparison pulls. It will not manufacture an aide in a county where you cannot hire one, and publishing coverage you cannot staff is worse than publishing nothing — you will accept referrals you have to hand back, which is how you lose a discharge planner permanently.

It will not beat a hospital-owned agency for that hospital's own discharges, and it will not overturn a narrow Medicare Advantage network. It also will not help if your intake process still takes six hours to say yes. The shortlist gets you looked at; your response time closes it.

Finally, be careful about what you publish. Do not put anything patient-identifiable on a public page, and make sure your marketing person understands that a stated referral-to-start-of-care time is a promise a case manager will hold you to.

Monday's job is one page. List every county and ZIP you actually staff as text, every payer by name including the specific Medicare Advantage plans, your levels of care, your field-staff languages, and your after-hours admission policy in one sentence. No pictures of maps. Then have somebody who does not work for you ask an assistant to find a hospice in your county, and see whether you come back.

Frequently asked questions

We get almost all our referrals from two hospitals. Does any of this matter to us?

It matters most to you, because concentration is the risk. When one of those hospitals is acquired or signs a preferred-provider arrangement, you will need referrals from families and physician offices in a hurry, and that is the channel where being readable decides whether you are on the list.

Is this just search engine optimisation with a new name?

It overlaps, but the goal is different. Traditional search wanted a person to click. This wants a set of facts to be extractable and comparable — ZIPs as a list, payers by name, levels of care in plain words. A page can rank well and still be useless to a comparison because everything on it is a picture or an adjective.

How do we know whether it is working?

Ask every new admission's family where they heard of you and record the answer in your intake screen, and separately watch your web chat and after-hours call volume. If assistants are recommending you, the tell is usually inbound contacts that arrive already knowing your service area and asking a very precise question.

Do we need to join some kind of directory?

Care Compare is already the public spine and you cannot opt out of it, so start by making sure your own site does not contradict it. Beyond that, your own pages are the asset. Directories come and go; the counties you staff and the payers you take are yours to publish.

A note on who answers when the shortlist calls

The last step of every version of this is a contact attempt, and a good share of them still arrive as a phone call or a web chat at an hour when the office is closed. CallSphere builds AI voice and chat agents that answer business phone lines and web chat 24/7, capture the caller and the referral details, and book the follow-up. It does not decide clinical eligibility and it does not accept a referral for you — but it does mean the 8 p.m. call from a daughter three states away reaches a conversation instead of a voicemail greeting.