About

ClinicalTrials.gov is the registry the U.S. National Institutes of Health runs. It is authoritative and it is complete, and it is also hard to read: a study record is a wall of regulatory fields written for the people filing them, not for the people trying to understand the trial.

So the first thing this site does is render the same record for the person reading it. That part is table stakes.

The second thing is the point. Trial information was built to be recorded, not discussed. A record tells you when a study started, who sponsors it and what it measures. It rarely tells you what someone actually wants to know:

• Is this study really still recruiting?
• Why was it terminated?
• When will the results be out?
• How does it compare with the other trial for the same condition?
• What does the investigator make of it?
• Who do I ask?

Those conversations do happen. They happen in email, in conference corridors, on investor calls, in patient groups, in replies nobody else ever sees. The answer exists — it just reaches one person and then disappears.

This site gives every trial a public page for them, anchored to its NCT number, so an answer given once is there for the next person who asks. It is why a question marked for a study team is mailed to the contact address the registry publishes, and why the reply lands in the open thread rather than in one inbox.

The registry deliberately does not offer a place to talk, and that is a reasonable decision for a registry to make. It leaves a gap all the same.

Any ClinicalTrials.gov study URL works here if you swap gov for gg:

clinicaltrials.gg/study/NCT04280705

That is the whole idea. Every link to a study that already exists — in a paper, an email, a forum post, a press release — becomes a link to a readable version of the same record by editing two characters. It also means the paths on this site are not ours to choose: they are whatever the registry uses.

Not every question needs someone to answer it. Many answers are already in the registry, spread across its records or across the versions of one record. Read together, the records say things no single one does, and these come from the records themselves, so a page stays current on its own:

• What changed on a study, and when. Each registry update is kept and named: which eligibility line, which site’s recruitment, which intervention or phase.
• Which studies test a drug. A registry search for pembrolizumab finds about 3,000 studies; about 1,100 of them test it. The rest mention it as earlier treatment, use it as the comparator, or give it beside the drug under test. Drug pages marked “Read study by study” have had every study the registry finds read for this.
• Where a condition’s studies are. Condition pages follow the registry’s own filing, brought up to date in full, and take in the studies it filed under no condition when their own listed conditions name one. The larger ones say step by step why their count differs from the registry’s search. A condition whose filing is mostly mistaken says so on its page, and a study names its main condition when it lists several.
• Who runs a study, and where. A company that appears under several names is one sponsor page, and study sites are gathered into hospital pages.
• How long studies like it take. A study’s dates are compared with similar studies of the same condition and phase.

These figures are ours, not the registry’s, and they can differ from it on purpose: the two counts answer different questions. How the figures are worked out says what goes into each one and what is left out.

Not affiliated with, endorsed by, or operated by the National Institutes of Health, the National Library of Medicine, or ClinicalTrials.gov. The similarity of the name is the mechanic described above, and it is exactly why this page says so plainly.

The registry remains the authority for every record. Where a record shown here differs from the one on ClinicalTrials.gov, the registry is right. A count or grouping worked out here can differ on purpose, as described above.

Not medical advice. A study appearing here is not a recommendation to enrol in it, and nothing written here — including by someone carrying a study team badge — is a substitute for talking to your own doctor.

ClinicalTrials.gg is built and run by Yongjang Jo, an individual, in Seongnam, Republic of Korea.

I spent four and a half years as a machine learning engineer in B2B healthcare, working with clinical trial data every day. What stayed with me is how closed the field is about information that is already public: the records are open, and it is still hard to find out what is actually happening with a given trial — whether it is moving, whether anyone can still join, what the people running it make of it. There was nowhere those questions could be asked in the open, so I built one.

I would like it to be useful to more than one kind of reader: to someone with a rare disease looking for a study they could join, to a clinician trying to explain one to a patient, to a researcher weighing their own work against what else is running, and to the sponsors and investors who need to see how a programme is moving. Those people rarely read the same page today. This is an attempt at one they could share.

You can reach me at contact@clinicaltrials.gg. Corrections to a record are best reported to ClinicalTrials.gov — they hold the data — but tell me too if this site is showing something the registry does not. If a name is grouped under the wrong company, use “Something wrong?” on the sponsor page. If a study is counted in the wrong role, signed-in readers can use “Report a problem” on the drug page.

• Every study record, synced with the registry every day — see where this data comes from.
• On each study page, a one-paragraph summary, its main condition, how it compares with similar studies, and what its dates show when read side by side.
• What changed on each study, by name: status, dates, sites, eligibility lines, interventions, phases, collaborators, documents and posted results, taken from the registry’s own updates. Recorded since September 25, 2026.
• Search by condition, drug, sponsor, hospital, status, phase, country, age and sex. Picking a suggestion as you type reaches every study filed under that condition, drug, company or country, not only those that mention the name.
• Drug pages that say how each study gives the drug, with the way from the registry’s count to the page’s. Dozens are read study by study, and a reader can ask for another to be.
• Condition pages with what is being tested, what reports within a year, broader and narrower conditions, the studies mainly about the condition, and, on the larger ones, why the count differs from the registry’s.
• Pages for sponsors, countries and hospitals, each listing its studies.
• A timeline on some study pages for now, showing when planned dates moved and how long similar studies take. Condition pages show those durations by phase.
• Up to five eligibility sets, kept in your browser, to filter by age and sex and see which studies fit. They compare age and sex only, which is not the same as being eligible.
• A checklist that goes through a study’s criteria line by line and gathers what to ask your doctor or the study team. Nothing is scored, and it stays in your browser.
• Finding a study’s sites by country and city, and an email to a site drafted in your own mail app.
• A discussion thread on every study and on drug pages read study by study, with moderation.
• Study team badges, verified against the contact address the registry publishes on that study’s own record. Everyone else is a member, with no title next to their name, on purpose: an unverified “sponsor” label would be worse than none.
• Following a study, with up to five of its sites watched, and a daily email saying what changed on its record, at those sites, within 50 km of them, or in its discussion.
• Saved searches, and an email on Saturdays when one of them has changed.
• Downloading your data from your account page, and turning each kind of email off on its own.

What changed lately is on What’s New.