The size of a bed tower is decided by the operating model, not the census forecast

A bed tower is presented to a board as a number. The number arrives with a demand study behind it: population growth, market share, discharges by service line, the patient days those discharges will produce. The forecast looks like the thing to argue about.

The forecast is only half the calculation. A patient day does not require a bed by itself. It requires a bed under a particular set of operating assumptions, and those assumptions decide the size of the building at least as much as demand does. Two hospitals with the same forecast can justify towers of different size, and both can be right.

Demand becomes beds only through assumptions

The conversion from patient days to beds rests on a handful of inputs. How long patients stay. What share of beds the hospital expects to have filled on an average day. How much the census swings by season, by weekday and by hour. Whether rooms are private or shared. How beds are grouped into units, and which units can take which patients.

None of these is a fact about the market. Each is a statement about how the hospital will operate in the year the tower opens and in the decades after. A forecast that is careful about demand and silent about these inputs has not sized the building. It has handed the sizing to whoever chose the defaults in the spreadsheet.

Length of stay is a capacity lever the hospital already owns

Average length of stay sits in the denominator of every bed calculation, and it is not fixed. Discharge planning, case management, weekend services, timely imaging and consults, and the availability of post-acute placements all move it. A hospital that shortens stays through operational change gains capacity without pouring concrete; a hospital that assumes today's length of stay will persist is building to accommodate today's delays.

That puts a direct question to the owner before design starts: which length of stay is the tower being built for? If the answer assumes improvement, the operational program that will deliver it needs a sponsor and a timeline, because the building will be sized as if it succeeds. If the answer assumes no change, the board is being asked to fund beds that partly house discharge delays, and should be told so.

The occupancy target is a statement of risk tolerance

A hospital cannot plan to fill every bed on an average day. The planned occupancy rate is the margin between average demand and the peaks the hospital intends to absorb without boarding patients in the emergency department, holding post-surgical patients in recovery, or diverting ambulances.

Choosing that rate is choosing how often the hospital will be full. A higher target builds fewer beds and accepts more days at or near capacity. A lower target builds more beds and accepts more of them standing empty on a quiet day. The right answer differs by service: a small specialty unit with unpredictable admissions needs more slack than a large medical unit whose census moves slowly. It is a clinical and financial judgment for the owner's leadership, not a convention carried over from another project.

Room policy changes how many beds are usable

Shared rooms carry a hidden capacity cost. A bed in a semi-private room is unavailable whenever the occupant of the other bed needs isolation, when patients cannot be paired, or when one occupant's condition makes a roommate unsuitable. On paper the room holds two patients; on a busy night it may hold one.

Facility guidelines adopted in many states favor single-patient rooms in new acute care construction; the design professionals and the authorities having jurisdiction determine what applies to a given project. The owner's decision sits upstream of that: a tower of private rooms behaves differently from the shared-room floors it may replace, and the bed count derived from the old building's experience needs to be adjusted for that difference rather than copied.

The forecast tells the owner how much care is coming. The operating model decides how much building that care requires.

Unit size is set by the staffing model

Units are staffed as wholes, with a charge nurse, a support structure and a span of observation the floor plate has to make workable. The number of beds on a unit is therefore tied to how the hospital intends to staff it, and the size of the tower follows from the number of units multiplied by the beds on each.

This is where a small change in assumptions becomes a large change in building. If the staffing model favors units of a certain size, a demand figure that falls between two multiples of that size forces a choice: build the extra beds, or run one unit undersized. The choice is operational before it is architectural, and it should be made by nursing and operations leadership with the cost of each option visible.

Flexibility is a decision to fund, not a hope to rely on

Acuity-adaptable rooms, universal beds that can step up or down in level of care, observation capacity separated from inpatient beds, and shelled floors held for a later fit-out all change how much capacity a tower provides, and when. Each has a cost. A room built to support critical care costs more than a room built for medical-surgical care, and a shelled floor carries structure, vertical transportation and infrastructure capacity before it carries a patient.

These options are worth having only if the operating model will use them. Universal rooms deliver flexibility only if staffing and admission policy let patients stay in place as their acuity changes. Shell space is useful only if the infrastructure to fit it out has been reserved and the funding path to complete it is understood. Without those, flexibility is a cost in the budget with no change to how the hospital runs.

Patient flow decides whether a bed is available

A tower adds beds at the end of a chain. Emergency department throughput, surgical scheduling, transfer acceptance, bed assignment and discharge timing determine whether a bed that exists is a bed that can be filled when a patient needs it. A hospital that discharges late in the day is short of beds every afternoon, however many it owns.

Some of the capacity a forecast appears to require can be found in that chain. The question for the owner is not whether to fix flow instead of building, which presents a false choice, but which flow assumptions the tower is sized on, and whether the hospital intends to change them.

What to settle before the tower is sized

Before a bed count goes into a design brief, the owner can require a short written statement of the operating model behind it: the length of stay assumed by service, the occupancy target and the reasoning for it, the room policy, the unit size and staffing model, the role of flexible and shelled capacity, and the flow assumptions. Each line should name the leader accountable for the operational change it depends on.

That statement does two things. It gives the design team a brief that will hold, rather than a number that will be reopened when operational leaders see the floor plans. And it gives the board something it can test: a set of decisions it can agree or disagree with, rather than an output it can only accept. Testing that basis before capital is committed is part of a readiness review and of independent review of the capital plan on a hospital project. A tower that opens with the wrong number of beds stays wrong for a long time, and the error can be traced to an assumption nobody was asked to own.