The Product
Evidence appraisal · rigor, not a recap
AI summarizes papers. TrialReviewer appraises them.
Not a summary tool — TrialReviewer appraises evidence like a clinical epidemiologist: recomputes absolute benefit ( / ), surfaces bias and design flaws, grades quality (), and states whether the evidence supports the claim. The decision stays yours.
Includes ACCORD, SPRINT, and RECOVERY examples
The ACCORD trial randomized 10,251 patients with type 2 diabetes at high CV risk to intensive glucose lowering (HbA1c<6.0%) or standard therapy… the intensive arm had higher all-cause mortality (HR 1.22, 95% CI 1.01–1.46) and was terminated early.
| Endpoint | Direction |
|---|---|
| Nonfatal MI | Favors intensive |
| CV death | Favors standard |
| All-cause mortality | 1.22 · favors standard |
| Severe hypoglycemia | ~3× excess |
Intensive control did not reduce MACE; all-cause mortality rose (HR 1.22) with ~3× more severe hypoglycemia — the evidence does not support tighter glucose targets in this population.
Every threshold judgment carries an inline [Author Year] citation, e.g. [Walsh et al. 2014].
Why not just use ChatGPT?
Same number, two very different treatments. Say a trial reports a 25% relative risk reduction:
“The trial reports a 25% relative risk reduction, suggesting the intervention may reduce the outcome compared with control.”
But it does not check absolute benefit, harms, or whether the result supports the clinical claim.
- Absolute risk reduction 1.6% (not 25%)
- Number needed to treat 61
- All-cause mortality actually increased
Appraisal: the evidence does not support the claimed benefit.
What you get
Reports follow the appraisal framework. Each dimension below maps to real engine output — not a fixed template with empty slots.
Detects study type (RCT / meta-analysis / observational) and flags the source — randomization, blinding, attrition, ITT, and more.
Eligibility and representativeness spelled out so you don't over-extrapolate.
and recomputed — not just a relative risk reduction (RRR) recap.
, , serious adverse events, and result fragility where relevant.
Component directions separated so a headline win can't hide harm.
Selective reporting, design-claim mismatch, immortal time bias — itemized with impact explained.
evidence quality and appraisal confidence stated separately — never conflated.
One Bottom Line sentence — whether the claim holds, stated plainly.
Why you can trust this appraisal
Only claims we can verify — we'd rather say less than oversell.
Every threshold judgment carries an inline [Author Year] citation, backed by a 60+ core reference library you can check in Expert reports.
Appraisals on real trials — ACCORD, SPRINT, RECOVERY — aligned with published expert commentary. Sample reports are the validation cases; compare yourself.
Six-layer evidence safety architecture + + — an explicit path, not a black-box score.
No clinical decisions for you, no fabricated numbers; when evidence is thin or out of scope, the report says so plainly.
The appraisal rests on a framework
Every appraisal is grounded in established evidence-appraisal methods and 60+ core references — not the model's intuition.
-informed rating — each downgrade criterion explained from a high-quality start.
appraisal confidence — a fatal methodology flaw sends the rating straight to D.
Practice-Changing Checklist — explicit criteria, not intuition.
/ / together — so relative-effect hype doesn't mislead you.
— how few events would flip statistical significance.
MCID — whether the effect clears a clinically meaningful threshold.
Sources: Nuovo 2002 · Walsh 2014 · Balshem 2011 · Schulz 2010, and 60+ more.
Two modes — pick how you read
Judge whether the evidence is strong enough to change practice — appraisal confidence and evidence quality kept separate, Concerns itemized and checkable, every threshold backed by an inline [Author Year] citation, full References at the end.
Read a published paper or diagnose your own draft. Teaching Mode walks through the methods, explains concepts, and surfaces the questions your teacher, journal club, or committee may ask.
Ready to appraise a trial?
Start with a sample report, or paste abstract/full text or upload a PDF.