Managing capacity, throughput, and the cost of an empty bed

Length of stay (LOS): the average number of days a patient occupies a bed. Hospitals watch the gap between actual LOS and GMLOS (Geometric Mean Length of Stay), the expected stay for a given DRG. When actual exceeds expected, the hospital is losing money on that fixed payment and, more importantly, blocking a bed someone else needs.

The three collisions on the whiteboard

Collision 1: ED boarding

Boarding means a patient has been admitted (the decision to keep them is made) but is still physically in the ED because no inpatient bed is ready. Those nine red magnets.

Boarding is expensive and dangerous. The ED cannot see new patients while its beds are full, so ambulances get diverted and the waiting room grows. Studies from groups like the Agency for Healthcare Research and Quality link crowding and boarding to worse outcomes and higher mortality risk. Financially, a diverted ambulance is a lost admission, and a lost admission is lost DRG revenue.

Collision 2: OR block scheduling

Surgeons are assigned block time: reserved OR hours, say every Tuesday morning. Elective surgery is the most profitable part of most hospitals, so protecting those blocks matters.

But surgery creates demand for inpatient beds after the procedure. If Monday's ED admissions fill every bed, tomorrow's scheduled knee and hip cases have nowhere to recover. Canceling an elective case wastes an expensive OR slot, angers the surgeon, and delays a paying patient. The OR director is not being difficult. She is protecting the hospital's margin engine.

Collision 3: discharge delays

The four "medically ready" patients waiting on a nursing home are the quiet killers of throughput. The clinical work is done. The bill is not growing in a good way. Each occupied bed blocks an ED admission upstream.

Discharge delays usually are not clinical. They are logistical: a skilled nursing facility (SNF) has not accepted the patient, home oxygen has not been delivered, a family meeting has not happened, or a physician has not signed the discharge order early enough.

How good hospitals unjam the flow

The huddle is not about assigning blame. It is about matching supply and demand for the next 24 hours. Here is what disciplined operations look like.

Discharge before noon

Most discharges cluster in the late afternoon, but most admissions (and ED surges) arrive earlier. That mismatch causes the afternoon bottleneck. Hospitals push for a target percentage of discharges before noon so beds open up before the demand peak.

Tactics: physicians round on likely-discharge patients first, pharmacy prepares medication lists the night before, and case managers start SNF placement on day one, not day three.

Predicting the discharge date on admission

Leading hospitals set an estimated discharge date (EDD) the moment a patient is admitted. It becomes a shared target. Everyone (nursing, therapy, social work) works backward from it. Without an EDD, no one owns the exit.

Smoothing the OR schedule

Counterintuitively, spreading elective surgeries more evenly across the week reduces peak bed demand more than any single discharge fix. Monday-heavy surgical schedules create a mid-week bed crunch. This idea comes from the work of the Institute for Healthcare Improvement on variability reduction: the biggest source of chaos is often the schedule the hospital controls itself, not random emergencies.

🎬 [VIDEO: "Improving Patient Flow" — youtube.com — an Institute for Healthcare Improvement overview of how variability and flow interact in hospitals]

A command center view

Larger systems now run a capacity command center: a room with live dashboards showing every bed, pending discharge, and inboundinboundA strategy that attracts prospects organically via valuable content (blog, SEO, social) rather than interrupting them.View full definition → transfer. Think air traffic control for patients. It replaces the phone-tag scramble with a single source of truth.

Putting numbers to the empty bed

Let's make the cost concrete with a simplified illustration. These are not real figures for any specific hospital, just arithmetic to show the mechanics.

Suppose a medical-surgical bed costs about 1,500 dollars per day to staff and run (an estimate for illustration only). A patient stays two days longer than their GMLOS. Under a fixed DRG payment, that is roughly 3,000 dollars of cost with no added revenue.

Now multiply the downstream effect: that blocked bed forced one ED boarding, which caused one ambulance diversion, which lost one admission worth, say, 12,000 dollars in DRG revenue. The two extra days did not just cost 3,000 dollars. They cost the opportunity of the next patient.

A simple way to frame the daily math:

Every morning huddle is really running this equation in people's heads. The best ones make it explicit on the board.

Where the incentives get tricky

Throughput improvement can collide with other priorities, and a fluent professional should see the tensions.

Quality versus speed: Pushing discharges too hard raises readmission risk (a patient bounces back within 30 days). Medicare penalizes excess readmissions through its Hospital Readmissions Reduction Program, so a hospital that games LOS by discharging too early can lose money twice: once on the penalty, once on the second unpaid stay. Fast and safe are not opposites, but speed without a solid discharge plan backfires.

Payer mix matters: Not every bed-day is equal. A commercially insured surgical patient and a Medicaid medical patient consume the same bed differently in margin terms. Hospitals must be careful here, both ethically and legally, since patients cannot be triaged by insurance. The point for operations is that the same throughput gain has different financial weight depending on who fills the bed.

Staffing is the real constraint: A "bed" is not a mattress. It is a staffed bed. If there are no nurses, the bed does not exist for capacity purposes. In tight labor markets, staffing, not physical space, is what caps occupancy. This is why nurse-to-patient ratios sit quietly behind every capacity conversation.

Reading the dashboard like an operator

When you walk into a hospital and hear these metrics, translate them fast:

• Occupancy climbing above the mid-80s: expect ED boarding within days.
• Actual LOS drifting above GMLOS: margin erosion and hidden capacity loss.
• Low percentage of morning discharges: a self-inflicted afternoon bottleneck.
• Rising readmission rate alongside falling LOS: the hospital may be discharging too aggressively.

None of these numbers means much alone. Together they tell you whether patients (and money) are flowing.

Key Takeaways

• Under fixed DRG payments, extra days cost money without adding revenue, so throughput, not occupancy, is the real financial lever.
• Above roughly 85 percent occupancy, hospitals lose the slack needed to absorb surges, and ED boarding follows.
• The three chronic bottlenecks (ED boarding, OR block scheduling, and discharge delays) are interconnected: fixing discharges upstream relieves both.
• Set an estimated discharge date at admission and target morning discharges to open beds before the daily demand peak.
• Speed must be balanced against readmission risk and staffing limits: a bed with no nurse is not capacity, and a rushed discharge can cost the hospital twice.