About and method
This site is educational history, not medical advice. Results from different trials are not directly comparable, and any comparison shown is illustrative. Talk to a clinician about decisions affecting your care.
How we handle evidence
- Every number is traced to a primary source (journal article, regulatory document or trial registry).
- If a value cannot be confirmed it is left out, or the entry is marked “needs review”. Nothing is approximated.
- Citations are checked automatically against PubMed, Crossref and ClinicalTrials.gov, and values are re-checked by a second reviewer.
- Results are shown as published. Comparisons are only drawn between trials in the same population and setting, and are illustrative.
Verification status
Each entry carries a status. Verified means a second reviewer, independent of whoever gathered the material, re-checked every number, name and date against the source. Needs review means something could not be confirmed (for example a paywalled full text), with a note saying what. Researchers never mark their own work as verified. Everything here was checked against primary sources to the extent possible; where that was not possible the entry says so.
How years and dates are chosen
A milestone sits at the year of its first publication online or its regulatory action, whichever came first. A reversal (a practice later shown not to help) sits at the year of the reversal itself. Where sources disagree the entry says so.
Curves and charts
Charts that draw a survival curve from reported medians or landmark values are labelled “Illustrative: modelled from reported values, not trial data” and can be switched off. Anything drawn before registry data exist is marked “Schematic: not data”. Cross-trial comparisons are shown only within the same population and setting and are never evidence that one treatment beats another.
Two reading modes
Explorer gives plain-language summaries with rounded figures. Clinician gives the reported estimates, intervals, endpoints and definitions. Both are views of the same verified record.
Known limits
- Guideline documents that could not be read in full are cited by identifier only and carry no extracted content.
- Some older papers have no abstract; their entries rely on the cited secondary sources and say so.
- Coverage is selective: the timelines show landmark steps, not every trial.
Report an error
If you find a mistake or a missing source, please tell us which entry and what the source says. Corrections are logged with a date. Email the error report.
Privacy and analytics
This site counts visits in aggregate with Vercel Web Analytics: pages viewed, the site a visitor came from, country and device type. It sets no cookies, does not follow visitors across other sites and keeps no personal profile. We collect your name or email only if you write to us.
Contact
Questions, collaboration, or a mistake to report: get in touch with Temirlan Kainazarov.