The Median Rare Disease Trial Recruits at Exactly One US Site

A family in Cheyenne, Wyoming and a family in Boston can have the same child, with the same ultra-rare diagnosis, and live in different worlds. The Boston family drives 20 minutes to a hospital running dozens of rare disease studies. The Wyoming family looks at the same ClinicalTrials.gov listing and starts pricing flights. Everyone who works in rare disease knows this gap exists. We wanted to know how wide it actually is.

So we measured it. We pulled every recruiting trial for all 195 rare diseases Trial Friend tracks, 6,372 unique studies in all, and mapped each one's listed study locations: 20,732 US site listings across all 50 states and Washington, DC. The resulting picture explains a great deal about why an estimated 90% of rare diseases still have no approved treatment, and why enrollment, not science, is so often the bottleneck.

More Than Half of US Rare Disease Trials Recruit at a Single Site

Start with the trials a US patient can theoretically reach: the 2,800 recruiting studies that list at least one American site. Their median US footprint is a single location. Not a network, and not a metro area. One named hospital or clinic, in one city, and if you do not live near it, the trial effectively is not recruiting you. In total, 1,522 of those trials (54%) recruit in exactly one US city, and when we tightened the definition to a single listed facility, the number barely moved: 1,457 trials (52%) name one building as their entire US presence. Meanwhile 2,035 (73%) run at 5 or fewer locations.

The average is 7.4 sites per trial, which sounds healthier until you notice what drives it: a small number of large Phase 3 programs with 50 or more locations pull the mean up while the typical study sits at 1. Averages flatter this system. Medians tell the truth about it.

The single-site pattern is not laziness. Rare disease trials are often run by the one academic center where the world expert works, funded on grants that cannot support 10 satellite sites, for a patient population that might number a few hundred nationwide. Understanding why it happens does not change what it means for the patient in the wrong state, though. Distance functions as an eligibility criterion that appears nowhere in the protocol.

Here is the number that surprised us most. Of the 6,372 recruiting trials across our tracked diseases, 3,572, a full 56%, list no US site at all. They recruit in Europe, China, Japan, and elsewhere, and for practical purposes they do not exist for an American patient. Rare disease research is genuinely global, and the US share of it is smaller than most American families assume when they search for options.

For 6 tracked diseases, the mismatch is total: every single recruiting trial sits outside the United States. Behcet's disease has 4 recruiting studies and not one American site. Thymidine kinase 2 deficiency, tyrosinemia type 1, aceruloplasminemia, heparin-induced thrombocytopenia, and anaplasmosis each have their only recruiting study abroad. An American family facing one of these conditions and hoping for a trial is, today, looking at international travel or nothing.

Where the Sites Are and Where They Are Not

The 20,732 US site listings are spread across every state, but spread is a generous word. California hosts 2,096 of them, with in-state recruiting options for 158 of our 195 diseases. Wyoming hosts 7 sites covering 3 diseases. Alaska has 21 sites covering 9. The median state offers in-state recruiting options for 72 of the 195 diseases, meaning that for the other 123, a resident of that median state must cross state lines to enroll in anything.

City concentration tells the same story from the other direction. New York City alone hosts 523 trial-site listings, Houston 502, and Boston 499, with Philadelphia, Chicago, and Los Angeles close behind. The top 10 cities account for 18% of every rare disease trial site in America. If you live in or near one of those metros, the system works reasonably well for you. Rare disease does not check zip codes before it strikes, and most Americans do not live in those 10 cities.

What Travel Distance Does to Real Families

External research puts weights on these distances. In CISCRP's Perceptions and Insights surveys, trial participants name travel logistics as the single top burden of participation, and in 2022, rare disease trial participants reported traveling an average of 135 miles to their site, double the 67 miles reported by participants in other trials (CISCRP, 2022). The EveryLife Foundation's landmark study put the total economic burden of rare disease in the US at $997 billion in 2019, a figure that includes exactly these non-medical costs: travel, lodging, and lost work days (EveryLife Foundation, 2021).

Distance also shapes who is in the data. When a trial's participants are disproportionately families who could afford to relocate for months or fly monthly, the evidence base itself skews wealthy, urban, and white. The FDA has been pushing trial sponsors toward decentralized designs, with home nursing visits, local labs, and telehealth check-ins, for precisely this reason.

“Decentralized clinical trials may enhance convenience for trial participants, reduce the burden on caregivers, expand access to more diverse populations, improve trial efficiencies, and facilitate research on rare diseases and diseases affecting populations with limited mobility.”

Robert M. Califf, M.D., FDA Commissioner, on the agency's decentralized trial guidance

The Diseases Where Geography Hurts Most

Beyond the 6 diseases with no US sites at all, 15 more have their entire US recruiting footprint inside 1 or 2 states. A few examples of what that looks like in practice:

Stiff person syndrome

2 recruiting trials, both in a single state. A patient anywhere else in America is traveling for every visit of a study that can run for years.

Wiskott-Aldrich syndrome

2 recruiting trials, 1 state. Gene therapy studies like these often require relocation near the center for a month or more.

Menkes disease

1 recruiting trial, 1 state, for a condition diagnosed in infancy where treatment timing is measured in weeks.

[Galactosemia](/trials/galactosemia)

2 recruiting trials, 1 state, for a metabolic disease detected on newborn screening in all 50 of them.

[Marfan syndrome](/trials/marfan-syndrome)

10 recruiting trials, which sounds like plenty, until you see the American sites span just 2 states.

The median tracked disease has recruiting sites in 18 states, which is workable for families near a site and still means long drives for everyone in the other 32. Only a handful of the most-studied conditions, the large cancers and multiple sclerosis, approach true national coverage.

What to Do With This If You Are Considering a Trial

First, check distance before eligibility. It sounds backwards, and it saves heartbreak: there is little point falling in love with a trial 4 states away before knowing whether closer options exist. The zip tool on our disease pages sorts this in seconds.

Second, ask every sponsor about travel support before ruling a trial out. Many rare disease sponsors reimburse flights, mileage, and hotels, and some arrange the logistics outright, precisely because they cannot fill a 40-patient study from one metro area. The site coordinator can tell you what is covered in one phone call.

Third, ask whether parts of the trial can happen near you. Post-FDA guidance, more protocols allow local labs, home nursing visits, and telehealth follow-ups, with travel reserved for dosing or imaging visits only. A trial that looks impossible as 26 round trips may be feasible as 6.

Methodology

We queried ClinicalTrials.gov's v2 API on September 12th, 2026 for recruiting-status studies against each of the 195 rare diseases Trial Friend tracks, using the same condition matching our live disease pages use, and deduplicated trials serving several diseases by NCT identifier. From each study we extracted listed locations, kept US sites, and counted one site per unique city and state pair per trial to avoid double-counting hospitals listed under 2 departments. As a robustness check we recounted at the individual facility level, which changed the single-site share from 54% to 52% and no headline conclusion. Site lists on ClinicalTrials.gov are sponsor-reported and can lag reality in both directions, and condition matching in a registry is imperfect, so treat exact figures as close estimates rather than gospel. The analysis script is re-runnable, and the full dataset behind this study is free to download and reuse from our open data page.