Questions about clinical trials, answered plainly

What is a clinical trial?

A clinical trial is a research study in which people volunteer to help test a medical approach under a written plan, called a protocol, that an independent ethics board (an institutional review board, or IRB) has reviewed. The approach might be a new drug, a new combination or dose of existing drugs, a type of surgery or radiation, a device, or a different way of delivering care.

A treatment trial tests a way of treating cancer, and trials ask different questions. Some compare a new approach with a standard treatment, or add it to one. Others, especially early-phase trials, look for a safe dose, study side effects, or check whether a treatment shows activity against a cancer, without comparing it to a standard treatment. When a placebo is used in a cancer treatment trial, it is often given alongside standard care rather than instead of it.

Taking part is voluntary. Participants sign an informed-consent form that explains the purpose, the schedule of visits and tests, the known risks, and the alternatives, and they can leave at any time. Clinical trials are how new cancer treatments are evaluated before approval, and for some people a trial offers access to a treatment that is not yet approved.

What do the phases mean?

Phases describe how far along a treatment is in testing, not how good it is. A later phase means more is known about it.

Many trials span two phases (“Phase 1/2”, “Phase 2/3”). Studies of procedures, devices, radiation techniques or behavioral programs are often “not phased”, because phases were defined for drugs.

How does a trial become available at a hospital or practice?

A trial begins with a sponsor — a pharmaceutical company, a National Cancer Institute cooperative group, a university, or a hospital — that writes the protocol and chooses where it will run. A hospital or practice becomes a site only after a good deal of work: a physician agrees to serve as investigator, research nurses or coordinators are assigned, the pharmacy and laboratory confirm they can handle the study drugs and tests, an ethics board approves the protocol, a contract and budget are signed, and staff are trained. Opening one trial at one site commonly takes months.

Sponsors choose sites differently. Companies tend to pick sites that can enroll quickly, which favors large centers in large cities. The National Cancer Institute runs networks built to spread trials more widely: the National Clinical Trials Network of cooperative groups, and the NCI Community Oncology Research Program (NCORP), which funds community hospitals and practices, including rural ones, to open federally sponsored trials. Each site reports whether it is open to the sponsor, and the sponsor updates the public record on ClinicalTrials.gov.

This is why access is uneven. Research staff and infrastructure cost money that the income from patient care rarely covers. A small practice can keep only a few trials open. Trials for rare cancers gather where those patients are referred. And a site that opens a trial may still enroll no one for months.

Why does “recruiting” differ from “available for me”?

Recruiting is a status in the public registry: the sponsor says the study is open and the site is enrolling. It is reported by the sponsor, it is not independently checked, and it can lag reality by months in either direction — a site may be listed as recruiting after it has filled its slots, or a new site may be open before the record catches up.

Available for me asks much more. A trial is written for a cancer type and often a subtype; for a stage or setting (before surgery, after surgery, metastatic, relapsed); frequently for a biomarker (a mutation, a receptor, a protein level); and for a number of prior treatments. It requires organ function and general fitness within set limits, excludes some other medical conditions, and may have only a few open places in the right group. Then someone has to think of it — many people join a trial because their own oncologist mentioned one — and the visits have to be possible.

Eligibility rules are often written narrowly so that results are clear and participants are kept safe, and professional groups and the Food and Drug Administration have published recommendations to broaden them where the narrowness is not justified. Even so, a county with twenty trials within reach may have none that fit one particular person, which is why this site also reports how many people have fewer than 20 trials within reach, an exploratory threshold, rather than only whether any trial is nearby.

What does this map measure?

It measures geographic availability: how many recruiting cancer treatment trials have a site within driving distance of where people live. Nothing more.

• Which trials. Recruiting, interventional, treatment-purpose studies on ClinicalTrials.gov with at least one US site, open to at least some adults aged 55 or older — on . Each study is tagged with the cancers it names. About three in ten name several cancers, or only general terms such as “solid tumors”, and form a multi-cancer group; a trial that names several cancers counts for each of them when you choose a cancer type.
• Where the trials are. Every recruiting site is placed at the center of its ZIP code (or at its city for the very large rural ZIP codes). A row that names a telemedicine or virtual service rather than a clinic is shown separately and not counted as a place.
• Where people live. Each of the country’s census tracts has a central point and a count of residents aged 55 and older. Distances run on the highway network from that point to each site.
• The figures. For each tract: the number of trials with a site within 60 road-miles. For a county, district, state or the nation: the share of residents 55+ whose tract has fewer than 20 (the map’s exploratory threshold), or none, weighted by population. Also the distance to the nearest NCI-designated cancer center and to the nearest county with 100 or more trials (a “broad menu”).
• Why 55 and older. Most cancers are diagnosed after 55, and the Census Bureau publishes this age group by tract, so the figures describe the people most likely to need a trial.
• Why 20 and 60. Both are exploratory thresholds chosen for this map, not clinical cut-offs. Twenty trials is a way of asking whether there is some choice nearby, since any one person qualifies for only some trials; this analysis has not established how many nearby trials a given patient needs. Sixty is measured in road-miles; how long that drive takes varies with roads, traffic and where people live. The Need-vs-access tab lets you change the access threshold.
• How often. The registry is re-read every Monday; every page says which pull it shows. Figures for a district or county can move when a single trial opens or closes.

The full method, every data source and the validation record are in Methods & data and on GitHub.

Which barriers does this map leave out?

Most of them. Distance is the barrier that can be measured for every community from public data; the others decide, person by person, whether a nearby trial is really available.

• Eligibility. The map counts trials, not trials a given person would qualify for.
• Whether a site is truly enrolling. The registry’s “recruiting” status is taken at its word.
• Repeated visits. Treatment trials often require visits every one to three weeks at first, plus extra scans and blood tests. A drive that is possible once can be impossible weekly.
• Cost and insurance. Medicare and, since January 2022, state Medicaid programs cover the routine care costs of qualifying trials, and most private plans must do the same for in-network care; but travel, lodging, parking, meals and lost wages are usually the participant’s own, and whether a site is in-network matters.
• Time, work and caregivers. Someone has to take time off, and often someone else has to drive.
• Transportation beyond distance. A household without a car faces a different problem from one with a long drive; the map’s household layers show where that is common.
• Referral. Many people join a trial because their own oncologist mentioned one. A practice that opens no trials may also refer to none.
• Language, culture and trust. Consent forms, study materials and staff may be available in one language only; past harms shape who is asked and who agrees.
• Telehealth. Some trials now allow some visits by video or at local clinics, which the map cannot see except for fully remote studies, which it lists separately.

The Your community page asks these questions alongside the local figures, and the What could help page says which of them each policy proposal would address.

What do these words mean?

NCT number

A study’s identifier on ClinicalTrials.gov, such as NCT01234567. Give it to a care team and they can find the study.

Sponsor

The organization responsible for a trial: a company, the National Cancer Institute, a cooperative group, a university or a hospital.

Site

A hospital or practice where a trial enrolls participants. One trial can have hundreds.

Eligibility criteria

The rules for who may join: cancer type and stage, biomarkers, prior treatments, organ function, other conditions, and sometimes age.

Biomarker

A measurable feature of a cancer or a person, such as a gene mutation or a protein level, that a trial may require or exclude.

Arm, cohort

A group within a trial that receives one treatment plan. Random assignment decides the arm in most Phase 3 trials.

Standard of care

The treatment that would be given outside a trial. New treatments are usually compared with it or added to it.

Informed consent

The process, and the document, by which a participant learns what a trial involves and agrees to join; it can be withdrawn at any time.

IRB

Institutional review board: the independent committee that reviews a protocol to protect participants.

NCI-designated cancer center

One of about 70 centers that treat adults and are recognized by the National Cancer Institute for their research; they run the widest range of trials, including early-phase studies.

NCORP

The NCI Community Oncology Research Program: a network that funds community hospitals and practices to run federally sponsored trials.

Basket or umbrella trial

A trial that enrolls several cancers sharing a target or biomarker (basket), or tests several treatments within one cancer (umbrella). They widen what counts as “a trial near me”.

Road-miles

Distance along roads, measured on the Census Bureau’s road network, rather than in a straight line.

Population-weighted

A figure for a county, district or state that counts people, not land: each census tract contributes in proportion to its residents 55 and older.

For research teams: planning where a trial runs

The same data that tells a family how far they are from a trial tells a research team, a cooperative group or a cancer center where its trials do and do not reach. Everything below is on the map or in the public repository.

• Filter to your portfolio. Filter trials narrows every figure by phase, sponsor class (industry, NIH/federal, academic and other), cancer type and patient age; the national, state and district figures are recomputed from census tracts in the browser.
• See who is out of reach. The default layer sizes each county by the number of residents 55+ with fewer than 20 trials within 60 road-miles; Need vs access sets cancer incidence or death rates beside access, county by county, with a downloadable table.
• Compare places on the same measures. Compare counties lists every county in a state; the Trials and Sites tabs list what is open in any place, with CSV downloads.
• Model a new site. A research prototype ranks candidate facilities by the residents 55+ a new site would bring within reach; it is under review and available on request.
• Reuse the data. County, district, state and tract tables, the road distances and every script are on GitHub; the registry is re-read weekly and each pull is archived. The methods and the validation record say what each figure can and cannot support.

Related policy and practice: FDA guidance on decentralized trial elements (September 2024) and on broadening eligibility criteria (December 2025), and the NCI Community Oncology Research Program, are summarized with official links on What could help.

Guided explorations

Each one opens the map on a view, keeps two or three prompts in a small box beside it, and lets you change every setting. They work for a teaching session, an advocacy meeting, or an hour on your own.