Straight answers to the questions people actually ask about clinical-trial risk and disclosure — the domain, not our product. Every figure below is a live measurement from our public facts feed, stamped with when it was measured and what it counted. Free to quote with attribution.
Public registry records carry early warning signs before a trial is formally stopped: a primary completion date that keeps slipping, a status change to suspended or terminated, enrolment targets revised downward, sites added late in the trial, and long gaps between record updates. None of these is conclusive alone — a slipped date can mean a protocol amendment rather than trouble — so the signal is the pattern across several of them, read against how that sponsor and those sites have behaved before.
Eichor's daily scan raised 3 enrolment-distress flags in the trailing week. (measured 2026-07-28, trailing 7 days — source)
A flag is a prompt to look, not a verdict. We publish how our own past calls turned out — confirmations and misses together — at eichor.com/track-record.
Only partly, and it is worth being precise about how much. Public registries do not publish per-site enrolment counts over time, so the quantity you would need in order to predict a rate is simply absent. What public data supports well is RANKING — ordering candidate sites by likely performance — not forecasting how many patients any one site will enrol. That ceiling is a property of the available data rather than the model, so any vendor claiming accurate per-site enrolment prediction from registry data alone is worth asking about. Real per-site accuracy requires the site's own CTMS or EDC records.
We state this because it is the question a careful buyer eventually asks, and because a limit discovered later costs more trust than one disclosed early. Method notes: provenance · how we built it.
FDAAA 801 is the United States law that requires responsible parties for applicable clinical trials to register them and to submit summary results to ClinicalTrials.gov, generally within one year of the trial's primary completion date. When results are not submitted on time the record stays on the public registry showing the obligation as unmet, which is visible to anyone who looks. Enforcement has historically been limited, so a large standing backlog of overdue records persists — which is why the backlog is measurable from public data.
Eichor's Overdue-Results Index currently stands at 984 trials. This is a standing backlog — the size of the outstanding set at a point in time — not a rate, and not a share of all trials. (measured 2026-07-28, current index size (standing backlog, not a rate) — source)
Primary source: ClinicalTrials.gov policy documentation. Our public view: eichor.com/fdaaa.
Article 37 of EU Regulation 536/2014 sets the European obligation to submit a summary of clinical trial results to the EU Clinical Trials Information System after a trial ends, with a shorter timeline for paediatric trials. The practical difference from the US regime is where the record lives and what is public: a sponsor running the same programme on both sides can be compliant in one system and overdue in the other, so a disclosure picture built only from ClinicalTrials.gov is incomplete for any sponsor with EU trials.
Primary source: EU Clinical Trials Information System. Our EU view: eichor.com/ctis.
From public evidence you can establish a site's track record rather than its promises: how many trials in the relevant indication it has actually run, whether those trials completed or stopped early, whether the principal investigator's specialty matches the condition under study, how long the site has been continuously active, and whether regulatory inspection history exists. Specialty match matters more than it sounds — a physician listed at a facility running an oncology trial may be a dermatologist, and counting them as oncology capacity is a common way to overstate a site's fit.
Site-level claim and verification flow: eichor.com/sites.
The main public sources are ClinicalTrials.gov for US and much international registration, the EU Clinical Trials Information System for European trials, FDA enforcement and inspection records, SEC filings for listed sponsors, NIH RePORTER for federally funded research, and CMS Open Payments for industry payments to physicians. What is absent matters as much: none of these publishes per-site enrolment counts over time, patient-level data, protocol amendment histories in full, or site-level operational metrics. Any claim requiring those either comes from a private data agreement or is an inference presented as a measurement.
Our daily scan indexed 54 newly registered trials in the trailing week — a measurement of what our scan indexed, not a registry-published statistic. (measured 2026-07-28, trailing 7 days — source)
Source-by-source detail: eichor.com/provenance.
Yes — the numbered facts on this page are published as a free, machine-readable feed with no API key required, and may be cited with attribution to Eichor. Each fact carries the timestamp it was measured, the window it covers, and a plain-language methodology note, so a reader can tell exactly what was counted. Values are recomputed on a schedule and the URLs are stable, so a citation keeps resolving.
Feed: eichor.com/facts.json · human view: eichor.com/facts · machine discovery: llms.txt. Suggested form: “according to Eichor (eichor.com/facts)”.
Answers are maintained by Eichor, a Biomedistats LLC product, from public evidence. Structured version for agents: /answers.json. Questions about Eichor itself are on the FAQ. If you believe an answer here is wrong, say so — hello@eichor.com.