Bed Management and Census for Residential Behavioral Health

How residential and detox programs run bed boards, midnight census, discharge projections and payer days, and what to test in an EHR with bed management.

Ask three people at a residential program how many beds are open tonight and you can get three answers. The house manager counts from the whiteboard. Admissions counts from a spreadsheet of expected discharges. Billing counts from last week's claims. Each number is defensible, none of them agree, and a detox bed sits empty for a day while a waitlisted caller goes somewhere else.

Ease is a platform for residential and detox operations, where census and billing run from the same system instead of a census board taped next to a separate biller. That is the test running through this guide: whether the bed board, the admissions pipeline and the claim read from one patient record, and where staff still step in by hand.

Below: bed boards and census counts, the admissions-to-bed handoff, occupancy math, reporting across sites, staffing, payer days, and a checklist for evaluating software.

What bed management and census mean in a residential program

Four terms get used loosely.

• Bed board: a live map of every bed at a location, showing who is in it, which beds are open, and which are out of service.
• Census: the count of patients present at a set point in time. Taking it at midnight lines it up with the way Medicare counts inpatient days, covered below.
• Occupancy: census divided by the beds available to fill, usually reported as a percentage over a week or a month.
• Projected availability: beds expected to open, based on each current patient's estimated discharge date.

These apply per location, at every overnight level of care from residential treatment to detox.

At least one state treats the daily census as a record to keep, with absences included. New Hampshire's licensing rule for substance use disorder residential treatment facilities (He-P 826.14(d)) has each licensee keep a system that identifies its daily census, "including times client is absent from the SUD-RTF." Other states may word this differently, so read your own rules. Either way, a patient out on a pass is still part of the census, and the absence needs a time on it.

How a census board works, from bed status to daily count

A residential census board is a hierarchy. An organization has locations, each location has houses or units, and each unit has beds. Every bed carries a status (occupied, open, being cleaned, out of service for repairs) and, when occupied, a patient and an estimated discharge date.

The daily count falls out of that structure when three habits hold:

1. Assignments change when the patient moves. A patient who steps down from detox to residential, or moves from one house to another, gets moved on the board the same day.
2. Discharges are entered when they happen. A discharge typed in two days late inflates the census and hides an open bed from admissions for two days.
3. Absences are recorded. Passes, hospital transfers and leaves of absence change who is physically present at midnight, even while the bed stays assigned.

In Ease, the bed board is organized by location, then housing unit, then bed, and staff assign and unassign patients there or from the patient's level-of-care tab. Each organization creates its own bed statuses, gives them colors, deletes the ones it no longer uses, and picks which status a bed takes by default when a patient is assigned or unassigned. Assigning and unassigning patients on the board is gated by ordinary role permissions, so who can change it is a deliberate choice.

The admissions-to-bed handoff

A referral comes in, admissions checks benefits and clinical fit, and the caller lands on a waitlist until a bed in the right level of care opens. The handoff from that list to a specific bed is where programs lose time.

Three questions make the handoff explicit:

• Who decides which bed? Admissions knows the caller. The clinical or residential lead knows the house, the roommate mix and the level of care. Name one owner for the final assignment.
• What triggers the call? A projected opening on the board is the usual trigger. Agree on how many days out admissions starts scheduling against it.
• What happens when the date slips? Estimated discharge dates move. If the clinical team extends a stay, admissions hears the same day, or a scheduled admission arrives to a full house.

A projected opening tells admissions when to start calling. It does not hold the bed. In Ease, the waitlist lives in the admissions CRM as a pipeline stage with configurable reasons, and intake runs as an ordered set of steps with handoffs and alerts between them. Nothing on the Ease bed board reserves or holds a bed for an incoming patient, so the bed is assigned once the person is admitted. Programs that pencil in a specific bed for a specific caller should decide where that note lives and who clears it.

Discharge projections, waitlists and occupancy math

The math is simple. The inputs are where it goes wrong.

Take a program with two houses and 48 beds, measured over a 30-day month:

• Available bed-days: 48 beds x 30 days = 1,440.
• Patient days counted at midnight: 1,296.
• Occupancy: 1,296 / 1,440 = 90 percent.
• Average daily census: 1,296 / 30 = 43.2 patients.

Two inputs can move that number while the houses stay exactly as full. If two beds were out of service for 10 days, available bed-days drop to 1,420 and the same patient days read as 91.3 percent. If each of the month's 40 discharges was entered a day late, the board counts 1,336 patient days and occupancy reads about 93 percent. The report looks better while admissions misses 40 open bed-days.

Projections work the same way. If 6 patients have estimated discharge dates in the next 7 days and 2 beds are already open, the program can plan for up to 8 open beds by next week, minus any stays the clinical team extends. Work the waitlist against that number by level of care, because a detox opening does nothing for a caller who needs residential. Good discharge planning is what keeps the estimated dates honest.

On the current Ease bed board, staff pick a future date and see how many beds are projected to be open, based on each occupant's estimated discharge date. Organizations still on the earlier Ease design see a simpler box: the number of beds projected to free up within a set number of days, which counts beds coming open and leaves out beds that are already empty. Either way, the projection is only as good as the discharge dates behind it.

Census reporting across facilities

Once an organization runs more than one site, the census question becomes how full each site and each level of care is, and the total, on the same date.

SAMHSA's facility survey asks for much the same cut: the 2026 N-SUMHSS client counts worksheet asks facilities for "the number of clients in treatment on March 31, 2026," breaks residential clients into detox, short-term and long-term treatment, and asks how many residential beds are designated for substance use treatment. The same worksheet suggests that organizations with several facilities make a copy for each one. A program that can only produce one blended number spends survey season rebuilding the split by hand.

For a COO, a useful weekly view has four columns per location: beds available, midnight census, occupancy, and projected openings over the next 7 days, with a total row across locations. Then reconcile: last week's census, plus admissions and transfers in, minus discharges and transfers out, equals this week's census. When it does not, a move or a discharge was missed.

In Ease, one organization holds many locations on one instance, each location has its own bed board, and census reporting is broken out by location. Admins configure which levels of care each site offers, so staff see only the options that apply at that site. Custom reports and intake-to-billing reporting sit in the same product, so census and billing figures draw on one patient record.

Staffing to census

Census drives the schedule. A detox unit at 4 patients and the same unit at 12 are different nursing assignments, and a residential house that fills on a Friday needs weekend coverage that was planned on Tuesday.

State licensing rules set the floor. New Hampshire's He-P 826 calls for enough qualified personnel to meet client needs at all times and at least one awake staff member on duty whenever clients are in the facility, and the same rule sets a fine of $500 per day for exceeding licensed capacity. Check your own state's staffing and capacity rules.

Above that floor, staff to next week's expected census as well as tonight's. Use the same projection admissions works from: current census, plus scheduled admissions, minus projected discharges, by day for the next two weeks. When a projected discharge slips, the staffing plan and the admissions plan move together.

Payer days and census accuracy

The census and the claim count days differently unless someone makes them match.

Medicare's convention for hospital and skilled nursing stays is a useful reference. The Medicare Benefit Policy Manual, Chapter 3 says "A day begins at midnight and ends 24 hours later." Under the same section, the day of admission counts as a full day, and the day of discharge does not unless admission and discharge fall on the same day. The day a leave of absence begins is not counted unless the patient is back by midnight.

Contracts with other payers for residential behavioral health care can define a billable day differently, including how they treat admission days, discharge days and passes. Write the convention for each contract into your billing policy, then make sure the census runs on the same clock.

The usual sources of drift between census days and billed days:

• Late discharge entry: the board shows a patient present after discharge, and the extra day is either billed and later denied or caught in review.
• Unrecorded passes: a patient on an overnight pass is still assigned to a bed but might not count as present at midnight under the contract.
• Level-of-care changes on the wrong date: a step-down from detox to residential entered a day late bills a day at the wrong rate.
• Transfers between locations: a move entered at the receiving site but never closed at the sending site counts one patient twice.
• Authorization limits: a day past the authorized stay can go unpaid without an extension, so compare the census with the authorization before the stay runs past it.

A weekly reconciliation catches this drift: midnight census days by patient and location, compared with billed days by patient and payer, with every difference assigned to a named person. When a stay is billed by the day, a one-day census error becomes a one-day billing error on every claim it touches. The UB-04 billing guide covers the institutional claim side in more detail.

In Ease, billing runs in the same product as the bed board. Ease generates institutional 837I claims and paper UB-04s with revenue-code-based service lines, and the biller reconciling a stay works from the same patient record that admissions and clinical staff use.

What to look for in a behavioral health EHR with bed management for residential programs

Answer these in a live demo, using your own scenarios.

1. Bed board by location. Can the vendor show every location, unit and bed, with statuses your team defines? Ask who can change an assignment and how that permission is set.
2. Moves in both places. Move a patient from detox to residential, and from one house to another. Do the level of care on the chart and the bed on the board change together?
3. Projected availability. Enter an estimated discharge date, then change it. Where does the projected opening show up, and does the count include beds that are already empty?
4. Waitlist to bed. Find a waitlisted caller and walk the path to an assigned bed. Ask whether anything holds a bed in advance, and if not, where staff record the plan.
5. Census across sites. Pull a midnight census for each location and the total for a past date. Then pull admissions and discharges for the same week and check that they reconcile.
6. Census to claim. Take one stay with a pass and a step-down. Show the census days, the billed days and the claim lines side by side.
7. One record. Count the logins and patient records a stay touches from first call to final claim.

Ease brings these pieces into one product: a per-location bed board with statuses your organization creates, projected availability from estimated discharge dates, a waitlist in the admissions CRM, census reporting by location, and institutional claims in the RCM, all under one login and one patient record. The limits deserve the same plain statement: the Ease bed board does not reserve beds, and the waitlist is a CRM stage rather than a bed-level queue. The residential treatment overview shows how the pieces fit a residential program.

Related reading:

• Signs You've Outgrown Your Behavioral Health EHR

Frequently Asked Questions

What is the difference between a bed board and a census?

A bed board is a live map of every bed at a location: who is in it, which beds are open, and which are out of service. A census is the count of patients present at a set time, often midnight, and it feeds occupancy, staffing and billing checks.

How do you calculate occupancy for a residential program?

Divide patient days by available bed-days for the period. A program with 48 beds over a 30-day month has 1,440 available bed-days. If it counted 1,296 patient days at midnight, occupancy is 90 percent and the average daily census is 43.2 patients.

Does a projected discharge date hold a bed for the next admission?

No. A projected opening is a planning signal built from estimated discharge dates, and the date can move. In Ease, the bed board shows projected availability, but nothing on it reserves or holds a bed, and the waitlist sits in the admissions CRM as a pipeline stage.

Why does the census not match billed days?

Usual causes are late discharge entry, unrecorded passes, level-of-care changes on the wrong date, transfers left open at the sending site, and different day-counting conventions. Medicare's manual counts days midnight to midnight and skips the discharge day unless admission was the same day, and other contracts set their own rules.

Can one EHR show census across several facilities?

It can when locations are built into the data model. In Ease, one organization holds many locations on one instance, each with its own bed board, and census reporting is broken out by location.