When the Stretcher Becomes the Bed — What ED Boarding Tells Us About a Reimbursement System Gone Wrong

A SAC perspective on the financial mechanics behind a public health crisis

In The Atlantic in April, Elisabeth Rosenthal described what emergency department boarding did to her husband Andrej during his treatment for esophageal cancer. In the summer of 2024, admitted to the hospital and needing an inpatient bed, he spent four days boarding in the emergency department before one became available. The experience left him with a standing refusal: “I will not go to the emergency room.” He died earlier this year. One of the physicians Rosenthal interviewed, Adrian Haimovich of Beth Israel Deaconess Medical Center, gave her the word the piece is remembered for. “Everyone knows about this problem, and no one cares enough to do anything about it,” he told her. “It’s barbaric.”

Anyone who has worked in or around hospitals knows this is not an outlier. Emergency department boarding is common, it is worsening, and it is dangerous. A Health Affairs analysis of 46 million hospitalizations from 2017 to 2024, co-authored by Haimovich, found that at the January 2022 peak, 40% of admitted patients boarded for more than four hours and 6.3% boarded for more than 24 hours. The Joint Commission has called it a public health crisis. The American College of Emergency Physicians has been raising the alarm for a decade.

The natural question is why.

It is not because hospitals are indifferent. It is not because capacity is being deliberately withheld. Emergency department boarding is the visible symptom of a deeper structural problem: the United States has built a hospital reimbursement system that does not pay for the care it requires hospitals to deliver. Over time, that mismatch has reduced the number of beds hospitals can afford to staff, until the system began to fail in plain sight.

This is the story behind the stretcher.

The numbers most patients never see

To understand how a capacity problem becomes a structural one, start with how hospitals are paid — specifically, what Medicare actually reimburses relative to the cost of care. The Medicare Payment Advisory Commission, MedPAC, is an independent body that advises Congress and has no incentive to overstate hospital distress. It found in its March 2026 Report to Congress that hospitals operated on a negative 12.1% margin on Medicare fee-for-service inpatient and outpatient services in fiscal year 2024, and projects the margin will remain deeply negative in 2026, at approximately negative 10%. In practical terms, for every $100 it cost hospitals to provide that care in 2024, Medicare reimbursed roughly $89.

Two qualifications belong with that number, and both come from MedPAC itself. The 2024 figure excludes $9 billion in one-time payments Medicare made that year to offset earlier underpayments for 340B drugs; had those been counted, MedPAC estimates the margin would have been negative 6.5 percent. And among the hospitals MedPAC identifies as “relatively efficient,” those that perform well on quality while keeping unit costs low, the median Medicare margin was negative 1 percent in 2024 and is projected to reach positive 1 percent in 2026. The shortfall is real and it is sustained. It is also narrower at the efficiency frontier, and 2024 looks worse than the underlying trend because of a one-time accounting event. For two decades running, Medicare has paid hospitals less than what their care costs.

The same report also explains why the capacity problem is easy to miss. MedPAC counts physical inpatient beds regardless of how often they are staffed, and on that measure capacity has been essentially flat: about 674,000 beds in fiscal 2024 against 675,000 in 2019, with aggregate occupancy near 71 percent. By that count, the beds are there. A bed that exists on a licensure form and a bed with a nurse assigned to it are not the same asset, and it is the second one that determines whether a patient in the emergency department has somewhere to go.

The population that fills those beds — older, sicker, often with multiple chronic conditions — is precisely the population whose care is reimbursed below cost. Behavioral health is even worse: AHA reported in its 2024 Costs of Caring analysis that inpatient psychiatric services were paid 34% below cost across all payers in 2023. That figure helps explain why psychiatric patients board in EDs for days at a time. The beds upstream have been closing for thirty years because no payer covers what they cost to operate.

How the cross-subsidy actually works

That gap does not remain contained on a balance sheet. Hospitals have to absorb it somewhere — and in practice, they do so by shifting costs onto the one set of payers that will negotiate: commercial insurers. The RAND Corporation’s most recent price transparency study found that commercial insurers paid an average of 254% of Medicare rates for the same hospital services at the same facilities in 2022. KFF found that hospitals in the top quartile of commercial payer mix had operating margins of 7.5%, compared to 3.3% for hospitals in the bottom quartile.

A hostile reader will see those numbers and conclude that hospitals are gouging private insurers. RAND’s own conclusion complicates both readings and belongs in any honest account of this. RAND found that most of the variation in commercial prices across hospitals is explained by hospital market power, and that very little is explained by a hospital’s share of Medicare or Medicaid patients. The hospitals charging commercial payers the most are not, in general, the hospitals absorbing the most public underpayment.

That finding is about why prices differ between hospitals rather than whether public underpayment shifts costs in aggregate, and both things can be true at once. But it makes the defensible version of this argument narrower than the familiar one. It is not that hospitals raise commercial rates because Medicare underpays. It is that every dollar Medicare and Medicaid fail to pay has to come from somewhere if a hospital is to keep its lights on, its nurses staffed, and its doors open at 2 a.m. when the ambulance arrives; that commercial payers are the only counterparty with room to negotiate; and that the hospitals with the least market power, disproportionately rural and safety net, have the least ability to do it. Those are the hospitals failing first, and that is the part of the cross-subsidy story the pricing data actually supports.

The cross-subsidy is also unraveling in two directions at once. Employer health premiums have risen steeply over the past decade, fueling political pressure to cut commercial rates. Meanwhile, hospitals serving populations with little commercial insurance — rural, urban safety-net, county systems — never had a meaningful cross-subsidy to begin with. The credit rating agencies tell a related story. S&P Global Ratings reported a 3.0-to-1 downgrade-to-upgrade ratio for nonprofit hospitals in the first half of 2024. HFMA, reporting on S&P’s full-year data, put the 2024 ratio at 4.5-to-1 against 3.8-to-1 in 2023, while noting that the absolute pace of downgrades slowed over the course of the year.

Boarding is not a billing strategy

This financial backdrop leads to a persistent misconception: that hospitals tolerate emergency department boarding because they get paid the same regardless of where a patient is treated. It is true that under Medicare’s diagnosis-related group payment system, the hospital receives a fixed payment per admission based on diagnosis, regardless of where the patient physically lies. So technically, yes, a boarded patient generates the same DRG payment as a patient in a proper bed.

But that framing collapses on the cost side. A 2024 study published in Annals of Emergency Medicine, using time-driven activity-based costing on acute stroke patients at an academic medical center, found that the daily cost of caring for a boarded medical/surgical patient was approximately $1,856, compared to $993 for the same patient in a proper inpatient bed. The study has real limits, being single site and single diagnosis at an academic stroke center, and the absolute dollar figures should be read with that in mind. One limit runs the other way. Those figures assume employed nurses only. When the study accounts for the hospital’s actual complement of traveler nurses, 35 percent in the emergency department and 13 percent on inpatient units, the daily cost of boarding rises to $2,258 against $1,095 for inpatient care.

The mechanism generalizes regardless. Boarding forces inpatient care into an emergency department workflow designed for rapid turnover, not sustained treatment, while simultaneously blocking access for new patients arriving by ambulance. Boarding does not zero out a Medicare DRG payment, but it erodes whatever margin existed and adds an opportunity cost the DRG never priced. The idea that boarding is a profit center is the inverse of the truth — hospitals that could discharge boarders into staffed inpatient beds tomorrow would do so tomorrow.

The reason boarding happens anyway is more sobering. It happens because the hospital has run out of staffed beds and has nowhere else to put a patient who cannot be discharged. Beds without nurses are furniture. Nurses without reimbursement that covers their wages are a permanent subsidy that hospitals running on thin margins simply cannot sustain.

What this means for hospital leadership

If boarding is not a choice and not a revenue strategy, then it is a constraint — and that constraint is now tightening in ways hospital leaders cannot ignore. For hospital executives, general counsel, and CFOs reading this, three points matter for the road ahead.

First, the regulatory environment is shifting in ways that will measure boarding without addressing its causes. In its CY 2026 Hospital Outpatient Prospective Payment System final rule, issued November 21, 2025, CMS finalized the Emergency Care Access and Timeliness (ECAT) electronic clinical quality measure. Voluntary reporting begins in 2027, mandatory reporting in 2028, and results factor into the 2030 payment determination through the Hospital Outpatient Quality Reporting Program. Hospitals that fail to meet OQR reporting requirements — which will include successful submission of the ECAT measure once it is mandatory — face a 2.0 percentage point reduction in their annual OPPS payment update. The consequence attaches to reporting compliance rather than absolute performance on the boarding measure, but public reporting of ECAT results creates its own reputational and contracting pressure. The measure assesses emergency department access, patient flow, and timeliness of care. Measurement is a precondition for any fix. It is also true that penalizing hospitals for a problem driven by chronic underpayment will accelerate, not solve, the capacity contraction. Hospital legal teams should be preparing comments and documenting the staffing and reimbursement constraints driving boarding times at their facilities now — that record will matter when the penalty phase arrives in 2030.

Second, throughput is being constrained from outside the hospital walls — and much of that pressure is litigable. Three patterns recur in our practice. Retrospective inpatient-to-observation downgrades under the two-midnight rule, where Medicare Advantage plans and commercial payers reclassify properly admitted patients weeks after discharge to convert a DRG payment into a far lower observation rate. Prior authorization delays for post-acute placement — skilled nursing, inpatient rehab, long-term acute care — that hold medically-ready patients in inpatient beds for days, which in turn keeps the ED full of admitted patients with nowhere to go. The HHS Office of Inspector General documented widespread inappropriate denials by Medicare Advantage plans in its April 2022 report, finding that 13% of prior authorization denials and 18% of payment denials in a sample from the 15 largest MAOs would have been approved under traditional Medicare. Post-acute facility stays — including transfers to skilled nursing and inpatient rehabilitation — were identified as one of the areas where inappropriate denials were especially concentrated. And 30-day readmission denials that recharacterize a clinically distinct second admission as a continuation of the first, eliminating payment entirely.

Update, June 2026: OIG has now completed and published two follow-on reports (OEI-09-24-00330 and OEI-09-24-00331). Reviewing prior authorization data from June 2024 across the 19 largest Medicare Advantage organizations, OIG found that MA plans denied 65% of prior authorization requests for long-term acute care hospitals, 54% of requests for inpatient rehabilitation facilities, and 12% of requests for skilled nursing facility admissions. When enrollees appealed skilled nursing facility denials, MAOs overturned 95% in favor of the enrollee — a rate OIG said raises concerns about the denials that were never appealed. OIG has said it will conduct an in-depth review of a sample of prior authorization case files in future work. The findings substantially reinforce the pattern this article describes: prior authorization operating as a delay and denial mechanism at the hospital-to-post-acute transition, with downstream boarding consequences upstream in the ED.

Each of these compresses inpatient throughput. Each shows up downstream as a patient on a stretcher in a hallway. Hospital revenue cycle, utilization management, and legal teams should be tracking these patterns at the claim level — they are the strongest evidentiary foundation when challenging payment denials and, increasingly, when negotiating MA contract terms at renewal.

Third, those pressures converge downstream, where the system has the least flexibility: discharge. The post-acute bottleneck is the real ceiling. Skilled nursing facility, inpatient rehab, and behavioral health bed capacity has contracted for years because reimbursement at those sites does not cover staffing. Hospitals cannot discharge patients to facilities that don’t exist. Boarding in the ED is the symptom most visible to the public, but the constriction begins three or four floors up — and increasingly, three or four facilities away.

The honest argument

Taken together, these dynamics point to a conclusion that is uncomfortable but difficult to avoid. It would be easier if the hospital industry were the villain in Rosenthal’s story. It is not. Hospitals are the institutions absorbing the consequences of a hospital reimbursement system that the public, the press, and policymakers have not yet decided to reckon with. When Rosenthal asks whether hospitals — “some of which are rich institutions” — should be required to open more beds, she is asking the right question of the wrong actor. The institutions that set Medicare and Medicaid rates, that approve Medicare Advantage business practices, and that have allowed inpatient psychiatric and post-acute capacity to collapse are upstream of every stretcher in every hallway.

Andrej Mrevlje deserved a real bed. So do the patients still boarding tonight. Giving them one will require CMS to revisit Medicare base rates that no longer cover the cost of care, state Medicaid agencies to confront rates that have not kept pace with inflation, and aggressive enforcement against MA plan practices that are converting hospital throughput into denied claims. Until those upstream actors move, hospital counsel can only do what counsel can do: defend the payments earned, document the constraints imposed, and build the record that will eventually force the conversation no one wants to have.

Sources provided upon request.