NewsResearch
AI Clinical-Trial Matching: Why a Suggested Match Is Only a Starting Point
AI may screen records against trial criteria. A research team must still confirm eligibility, location, status, risks, and practical fit.
Original commentary from the Cancer Explained editorial team.
Please note: this page is educational only — it is not medical advice, and it does not speculate about anyone’s health beyond reliable public reporting. For questions about your own health, talk with your healthcare team.
What a trial match actually requires
Trial listings are long, and the rules inside them are detailed. Software that reads those rules quickly is genuinely useful. The question is what a computer match proves.
NCI is direct about the process. Every trial has eligibility criteria, which are the requirements a person must meet to join. NCI's examples include having a certain type or stage of cancer, having had or not had a particular treatment before, having specific genetic changes in the tumor, being in a certain age group, medical history, and current health status.
Those facts live across scans, pathology reports, laboratory values, and treatment notes. A study team reads them and decides. That step does not move to software.
This page explains federal sources. It is not medical advice and does not suggest a test or treatment.
Why these criteria exist
They are not paperwork for its own sake. NCI explains that eligibility rules reduce medical differences among participants, lower the risk of harm, and keep the group to people most likely to benefit. When participants are alike in specific ways, researchers can be more confident the results came from the treatment and not from something else.
That is also why matching is hard. The rules are specific on purpose.
The criteria themselves are under review
Here is a point most coverage misses. NCI researchers have written that eligibility criteria have not kept pace with how trials are run. Restrictive rules have been a real hurdle for people who wanted to join, and they have also limited how far study findings can be applied to everyone else.
NCI has worked to broaden criteria on some of the trials it funds, including easing upper age limits in adult studies. So a matching tool is reading a rulebook that is itself being revised. A tool trained on old wording can carry old exclusions forward without anyone noticing.
A long list is not a good result
Counting suggested studies is the wrong measure. NCI's own guidance points to what happens next: check the trial locations, look closely at the protocol summary, and decide whether the site is manageable for you. Then contact the trial team, or ask someone on your health care team to contact them.
A useful evaluation of a matching tool would follow that same path. It would compare AI-assisted matching with the current referral process, and it would count what actually happened. Suitable matches confirmed by a study team. Contacts made. Screening visits. Enrollments. It should also test whether the tool narrows or widens the gap for people who live far from major cancer centers.
What a match does not settle
- A computer match is not confirmation of eligibility.
- A listed trial may not be recruiting, or may not be reachable from where you live.
- A close biological fit does not show that the experimental treatment will help.
- The trial phase shapes what the study can even find out. Early phases mainly test safety and dose. Later phases compare against standard treatment. Our page on the phases of clinical trials explains the difference.
Before you act on a list of trials
NCI's advice starts with a conversation, not a search box. Talk with your doctor or another member of your care team first. NCI's Cancer Information Service will also run a tailored trial search you can bring to that conversation, free of charge, at 1-800-4-CANCER. NCI notes that search results do not replace advice from your doctor.
If you are weighing a study, our guides to what clinical trials are and the risks and questions involved in joining one cover what to ask.
Questions to bring to that conversation
- Which eligibility facts did the tool confirm, and which did it assume?
- Is the trial actively recruiting at a site I can reach?
- What standard treatment am I comparing this against?
- What will travel, time off work, and caregiving cost me?
How this article was prepared
An AI-assisted editorial system helped prepare this page. No named medical reviewer has reviewed it unless one is listed.
The National Cancer Information Foundation publishes Cancer Explained. This page is for learning. It is not medical advice and does not suggest a test or treatment.
See an error, old source, or unclear wording? Tell us.
Know someone who needs this?
Plenty of people are looking for something like this and do not know where to start. If this would help a friend or someone you love, send it on — we have written an opening line so you do not have to stare at an empty message. You can change every word of it.
Your message is written and sent in your own email or messaging app — we never see who you send it to, and nothing is added to any list.
Put the story in context
Prevention, possible warning signs, screening, and diagnosis
This story relates to AI clinical-trial matching. The information below is general: it does not reveal anything else about a public person’s health, and not every point applies to every cancer. Personal advice depends on age, symptoms, family history, exposures, and medical history.
Prevention and risk reduction
Not every cancer can be prevented. Avoiding tobacco, protecting skin from ultraviolet radiation, limiting alcohol, staying active, and receiving recommended HPV or hepatitis B vaccination can lower the risk of certain cancers. A risk factor is not a prediction or a cause in one individual.
Symptoms and possible early signs
Possible signs vary and are often caused by conditions other than cancer. Changes worth discussing include a new lump, unexplained bleeding or weight loss, a persistent cough, lasting bowel or bladder changes, a changing skin spot, or symptoms that persist or worsen. Some early cancers cause no symptoms.
Screening and early detection
Screening looks for certain cancers before symptoms begin. Recommended tests exist only for some cancers and depend on age and risk. Screening can have benefits and harms; it is not the same as evaluating a new symptom, and there is no single routine scan or blood test that reliably screens for every cancer.
How cancer is diagnosed
Diagnosis may involve a history and exam, imaging, laboratory tests, and often a biopsy. Pathology can identify the cancer type and may test biomarkers that guide treatment. Symptoms, screening results, tumor markers, or online stories alone cannot confirm cancer.
A public story may encourage questions, but it should not be used to estimate your risk or choose testing. Contact a healthcare professional about a persistent or concerning change. Seek urgent care for severe or rapidly worsening symptoms.