When hospital length of stay begins to rise, attention often turns immediately to the emergency department. The ED is the front door, the pressure is visible, and its leaders are often the first to raise concerns about boarding, admission delays, and patients leaving without being seen.
Yet length-of-stay unplanned increases frequently begin farther downstream. Patients may have completed the clinical portion of their stay but remain in beds while teams coordinate post-acute placement, transportation, prior authorization, durable medical equipment, medication needs, or family arrangements. Across a hospital, those delays accumulate into blocked med-surg beds, delayed ICU step-downs, longer ED waits, and avoidable inpatient days.
Hospitals’ capacity margin is already narrowing. For example, a 2025 JAMA Network Open study found that average U.S. hospital occupancy reached 75.3% in 2024, compared with 63.9% before the pandemic. If staffed-bed supply and hospitalization rates remain unchanged, adult hospital occupancy could reach 85% as early as 2032.
Three signals can help leaders determine whether a throughput problem is rooted in discharge: a persistent gap between actual and geometric length of stay, a pattern of discharges drifting later into the day, and declining responsiveness across the post-acute network. Together, these signals show where time is being lost after treatment is complete.
Start with the gap that earns executive attention
The most useful place to begin is the difference between actual length of stay and geometric length of stay, or GLOS. Geometric length of stay reflects the expected stay associated with a patient’s diagnosis-related group and provides an evidence-based reference point for how long similar patients typically require inpatient care.
A single patient who stays longer than expected may have complex clinical circumstances, caregiver limitations, or unusual social needs. A consistent delta across a service line, unit, or patient population is the more meaningful signal.
Consider a diagnosis-related group with an expected stay of three days. If a hospital’s average actual stay for those patients is 4.5 days, the organization is carrying an additional day and a half beyond the expected benchmark. That variance affects capacity and financial performance because reimbursement generally reflects the expected course of care.
This metric also creates a common language for leaders who may otherwise view throughput through different lenses. Nursing sees workload and delayed transfers. Care management sees placement barriers and authorization delays. Physicians see clinical readiness and variations in practice. Finance sees the cost of avoidable days.
A persistent actual-to-expected gap gives these groups a shared problem to solve and helps secure executive support for cross-functional action. Leaders should pull this data before launching another broad throughput initiative, then examine where the variance is concentrated and which discharge barriers appear most often.
Late discharges reveal where planning breaks down
Discharge timing provides the next important signal. A patient who leaves at 5 p.m. rather than 11 a.m. may appear to represent a six-hour delay. The impact can extend much further when another level of care or services at home are required.
Skilled nursing facilities, inpatient rehabilitation facilities, and transportation providers often have daily cutoff times. Durable medical equipment suppliers have operating windows as well. If a hospital bed, oxygen, or another essential item is ordered late in the afternoon, delivery may no longer be possible that day. The patient then remains overnight because a logistical task missed its window.
The consequences move backward through the hospital. A med-surg bed stays occupied. A patient who is ready to leave the ICU cannot step down. The ICU bed remains unavailable for another critically ill patient. Admitted patients continue boarding in the ED, and wait times rise for everyone arriving behind them.
A 2025 Health Affairs study encompassing 46 million ED visits at 1,500 hospitals illustrates the scale of boarding across the nation. The study found that more than 25% of admitted ED patients waited at least four hours for an inpatient bed during non-peak months, with the share approaching 35% during winter.
Most hospitals already have discharge milestones and target times. Chronic late discharges should prompt an audit of whether the work begins early enough to meet them.
Discharge planning should become a day-before discipline. Teams can identify the next day’s likely discharges, confirm outstanding clinical milestones, place equipment orders, address transportation needs, begin authorization work, and engage families before the morning rush. This allows nurses, physicians, therapists, and care managers to start the day executing a plan instead of waiting for rounds or huddles.
The same discipline should begin at admission. Patients and families benefit from understanding the milestones required for a safe discharge, such as walking a defined distance without assistance, maintaining adequate oxygenation, or demonstrating medication readiness. When progress falls behind, the care team can intervene earlier rather than discovering the problem on the expected day of discharge.
Post-acute responsiveness is an operational lever
A third signal appears in the response rates of post-acute partners. Slow replies, repeated denials, and unclear acceptance criteria can translate directly into additional inpatient days. Although hospitals do not control every post-acute decision, they can improve how they manage and use their networks.
Health systems should track which facilities respond promptly, which accept the greatest share of appropriate referrals, what types of patients they can support, and how patients fare after transition. Regular conversations with these partners can clarify capacity, clinical capabilities, documentation requirements, and recurring reasons for denial. Partnering with a referral management platform that holds the network to a high-performance standard is another pathway to improving response rates.
Over time, this information can support a high-performing network of responsive partners with demonstrated outcomes. It also gives patients and families better guidance. Recommendations grounded in responsiveness and outcomes can speed placement and support confidence in the next site of care.
Technology can strengthen each of these practices when it creates shared visibility. Predictive discharge readiness can flag likely discharges earlier, surface missed milestones, and help every shift work from the same plan. That shared view matters in a 24-hour environment where teams change hands repeatedly and critical details can be lost in scattered notes, verbal updates, or multiple systems.
Reducing “chart surfing” also has value. Busy clinicians may lack the time to search for every relevant detail. Bringing forward mismatched therapy recommendations, unresolved equipment needs, or stalled referrals can reveal barriers that might otherwise remain hidden until discharge day.
Building resilience by reading the signals
Length-of-stay creep should prompt leaders to look closely at how patients leave the hospital. The actual-versus-geometric gap shows where avoidable days may be accumulating, discharge timing reveals whether planning begins early enough, and post-acute response rates show where external coordination needs attention.
Hospitals that learn to read these signals can move from reacting to congestion toward preventing it. With earlier planning, stronger partner visibility, and a shared view of discharge readiness, health systems can create more reliable patient movement and build a more resilient model of care for the years ahead.
About Lane Wise
Lane Wise joined ABOUT Healthcare in 2019 as Vice President of Clinical, bringing with him over 20 years of clinical experience in healthcare operations and patient care. Lane is a proven healthcare leader and has been successful in delivering clinical consulting and ongoing improvement initiatives with customers. Before joining ABOUT Healthcare, Lane was Director of Patient Navigation at PHI Health. Prior to that, he held nursing leadership and paramedic positions at West Tennessee Healthcare.
