When a patient arrives critically ill at a small hospital that cannot provide the care they need, the patient is moved. Inter-hospital transfer is one of the quiet, constant logistics of American medicine: hundreds of thousands of times a year, someone is loaded into an ambulance or a helicopter and driven or flown to a larger center. Transfer is often the right decision and sometimes a life-saving one. It is also slow, expensive, clinically risky, and frequently avoidable. This profile sets out the scale, the cost, and the timing of transfers, with one caution stated up front and kept throughout: transferred patients are sicker to begin with, so raw comparisons of their outcomes are not, by themselves, evidence that transfer caused harm. All figures are drawn from public data, listed at the end.
A constant of small-hospital care
For a small or rural hospital, transferring out is routine rather than exceptional. Rural emergency departments transfer their patients at roughly three times the rate of urban ones, about 6.2 percent of visits against 2.0 percent (Greenwood-Ericksen and colleagues, 2021). The reason is structural: the smaller hospital simply does not have the specialist, the operating room, or the intensive-care capability the case requires, so the patient has to go to where those things are.
The clock does not stop for the drive
For the most time-sensitive conditions, the transfer itself becomes part of the delay. The interval a hospital takes to move a patient out, the door-in-door-out time, is supposed to be brief. In practice it is not. For heart-attack patients needing catheterization elsewhere, the median door-in-door-out time was about 68 minutes against a 30-minute goal, and only about 11 percent of hospitals met that goal (Wang and colleagues, 2011). For stroke patients eligible for clot retrieval, the median was about 174 minutes against a 120-minute guideline, with only about 27 percent meeting it (Stamm and colleagues, 2023). Every one of those minutes is brain or heart muscle.
Every minute spent arranging a move the patient never needed is delay by another name — and for a failing heart or a stroke, that clock is already running.
How many transfers were necessary?
A substantial share of transfers turn out, in hindsight, to have been potentially avoidable: the receiving center discharges the patient quickly, or provides care the sending hospital could have given with the right support. Estimates cluster around one in four, though they range widely, from about 11 percent under the strictest definition to nearly half in a broad trauma sample (multiple studies, 2021 to 2024). The range is driven mostly by how avoidability is defined, not by genuine disagreement about the underlying reality, which is that a meaningful fraction of these costly, risky journeys did not have to happen.
The cost of the journey
Transfers are expensive, and the most expensive are by air. The median charge for an air-ambulance transfer was roughly 36,000 dollars by helicopter and 41,000 by fixed-wing aircraft, and about 69 percent of these flights were out-of-network, historically exposing patients to large surprise bills (Government Accountability Office, 2019). The No Surprises Act, effective in 2022, now bars balance-billing patients for air-ambulance transport, which protects the patient but does not reduce the underlying cost of moving people who might not have needed to move.
A word on outcomes, carefully
It is tempting to point to studies showing that transferred patients die more often or stay longer, and several do. But those comparisons are confounded at the root: hospitals transfer their sickest patients, so transferred groups start out worse, and much of the apparent penalty disappears once severity is accounted for. In some settings transfer is clearly protective, as when a severely injured patient is moved from a hospital that cannot treat them to a trauma center that can. The defensible claim is narrower and more useful: the harm worth chasing lives in the avoidable transfer and the delayed one, which together account for most of the cost and nearly all of the waste.
Keeping the patient and bringing the specialist
That is the gap telemedicine can close. If a remote specialist can assess a patient at the sending hospital in real time, some transfers become unnecessary, because the local team can safely keep and treat the patient with expert guidance. The evidence is encouraging but genuinely mixed, and it is worth being honest about that. Some programs show clear reductions: a national tele-stroke program cut emergency transfers sharply, and a tele-ICU program lowered ICU transfer rates. Others have reported the opposite, with transfers rising after telemedicine was introduced, sometimes because remote review surfaces patients who genuinely should be moved.
What telemedicine changes is the quality of the decision. By putting specialist judgment at the bedside before the ambulance is called, it lets each case be sorted on its merits. A transfer that happens because expert assessment confirmed it was needed is a good transfer; one that happens only because no specialist was on hand to weigh in is exactly the kind the system can no longer afford. The goal is a system that moves the right patients, on time, and keeps the rest safely where they are.
Sources and method. Rural and urban transfer rates: Greenwood-Ericksen and colleagues, JAMA Network Open (2021), Medicare data. Door-in-door-out times: Wang and colleagues, JAMA (2011) for heart attack, labelled as dated, and Stamm and colleagues, JAMA (2023) for stroke; the 30-minute and 120-minute figures are condition-specific benchmarks and are not interchangeable. Potentially-avoidable-transfer estimates: Li (2024), Van Schaik (2022), Wright (2021), Teng (2021), and May (2024); the wide range reflects differing definitions. Air-ambulance charges and out-of-network share: U.S. Government Accountability Office (2019), using 2017 data; the No Surprises Act took effect in 2022. Outcome confounding: transferred patients are systematically sicker than direct admits, and severity adjustment narrows or removes raw outcome gaps; transfer is protective in some settings, notably trauma. Telemedicine effects on transfer rates: Department of Veterans Affairs tele-stroke (2022) and tele-ICU (2018) programs, and a Mayo Clinic tele-ICU analysis (2017); effects differ by program and are reported here without vendor-funded sources. Where a figure is dated, confounded, or contested, the text says so.
