Causaly for clinical research

Advance your research, with stronger evidence

From identifying the unanswered questions to building study rationales, protocols and publications, Causaly helps you find, synthesize and verify scientific evidence in a fraction of the time.

Your research is never used to train models. ISO 27001 certified.

Research
Which randomized trials have tested colchicine after a heart attack, and where do they disagree?
SearchingPubMedFull-text papersClinicalTrials.govmedRxiv preprints
TrialPatientsPrimary endpoint, HR (95% CI)
COLCOT 20194,7450.77 (0.61–0.96)1
LoDoCo2 20205,5220.69 (0.57–0.83)2
CLEAR SYNERGY 20247,0620.99 (0.85–1.16)3
Source 1 · PubMed
Efficacy and Safety of Low-Dose Colchicine after Myocardial Infarction
N Engl J Med · 2019 · Randomized controlled trial

The primary end point occurred in 5.5% of the patients in the colchicine group, as compared with 7.1% of those in the placebo group (hazard ratio, 0.77; 95% confidence interval [CI], 0.61 to 0.96; P = 0.02).

Trusted by research teams at leading biopharma companies. Now open to clinician-researchers.

From the first idea to the protocol

Four questions every study has to answer

A clinical research fellow shaping a first fellowship application, a hospital research team drafting a protocol, a guideline panel weighing the trials: each has to show reviewers what is already known before anyone funds, approves or follows the work. That case rests on four questions, and a reviewer will check every answer.

Below, one real study idea taken through all four: low-dose colchicine after a heart attack.

  1. 1Is the question still open?What has been published, what is registered and running, and where nothing exists yet.
  2. 2Why is the study needed?Where earlier trials disagree, and why: the case for a new one.
  3. 3How should it be designed?The doses, endpoints, event rates and safety signals earlier trials used.
  4. 4Will the references hold up?Every claim in your draft, one click from the sentence it came from.
Question 1 of 4 · Is the question still open?

Know where the gap is before you write

Published trials, registered and ongoing trials, and the populations no one has tested yet, in one answer. Funders ask for exactly this.

The gap: no completed trial restricted to adults over 75, and none in chronic kidney disease or dialysis, where only trials still running are registered.

Which populations have randomized trials of colchicine for cardiovascular prevention?
Population Completed trials Registered, ongoing
Acute MI or ACSCOLCOT 4,745 · CLEAR SYNERGY 7,062TACTIC, 6,574 planned
Chronic coronary diseaseLoDoCo2 5,522 · LoDoCo 532Three post-PCI trials, up to 8,862
Ischemic stroke or TIACONVINCE 3,154 · CHANCE-3 8,343CHANCE-3 EX, 7,500 planned
Heart failureCOLICA 278, biomarker endpoint2,500 planned, ischemic HF
Type 2 diabetesSubgroups only, e.g. LoDoCo210,000 planned, factorial
Chronic kidney diseaseNone foundColchiRen, 744 planned
DialysisNone foundFeasibility study, 100
Adults over 75Subgroups onlyNone found
Dashed: no randomized trial found
Start your free trialCheck whether your own question is still open.
Next question2Why is the study needed?
Question 2 of 4 · Why is the study needed?

Make the case: where the trials disagree, and why

Conflicting trials are the reason a new one is needed. Causaly sets them side by side and explains the differences in population, timing and dose, each with its source.

TrialPatientsHazard ratio, primary endpoint (95% CI)
Why CLEAR SYNERGY differs

Started within 72 hours of PCI, in a mostly STEMI population. Commentators suggest the hyper-acute phase may respond differently from the weeks after MI, where COLCOT saw benefit.

● Laudani et al., 2025

Effect estimates as reported by each trial for its own primary endpoint. Endpoints differ between trials; this is not a meta-analysis.

Question 3 of 4 · How should it be designed?

Design it on what earlier trials did

The doses, timing, endpoints, event rates and safety signals of every earlier trial, side by side, so each choice in your protocol starts from evidence.

  1. The doses, timing and endpoints earlier trials chose, side by side.

Research
What are you exploring today?
Which doses, start times, comparators and primary endpoints have colchicine cardiovascular trials used?
SearchedClinicalTrials.govPubMedFull-text papersmedRxiv preprints
TrialRegimenStartedPrimary endpoint
COLCOT0.5 mg daily vs placeboWithin 30 days of MI (mean 13.5 days)CV death, arrest, MI, stroke, urgent angina revascularization1
CLEAR SYNERGYColchicine vs placebo, 2×2 with spironolactoneWithin 72 h of PCICV death, MI, stroke, ischemia-driven revascularization2
COPS0.5 mg twice daily for a month, then dailyDuring the ACS admissionDeath, ACS, urgent revascularization, stroke3
LoDoCo-MI0.5 mg daily vs placeboWithin 7 days of MIhs-CRP ≥2 mg/L at 30 days4
Question 4 of 4 · Will the references hold up?

Every claim, one click from its sentence

Before a reviewer, a co-author or an ethics committee asks where a number came from. Every claim opens on the passage it rests on, every search is on record, and the references export with their full metadata.

Your background section

In randomized trials, low-dose colchicine reduced cardiovascular events after a recent myocardial infarction1 and in chronic coronary disease2, but the largest trial, started within 72 hours of PCI, found no reduction3. Whether timing explains the difference is unresolved.

Source 1 · PubMed
Efficacy and Safety of Low-Dose Colchicine after Myocardial Infarction
N Engl J Med · 2019 · Randomized controlled trial

The primary end point occurred in 5.5% of the patients in the colchicine group, as compared with 7.1% of those in the placebo group (hazard ratio, 0.77; 95% confidence interval [CI], 0.61 to 0.96; P = 0.02).

  • Every search on recordRerun any search to see its full result set
  • Screened and referencedSee every document it read, and the ones it cited
  • References exportedAuthors, journal, dates, publication type and MeSH terms
Start your free trialCheck the references in your next draft.
NextTry a question in your field

Pick a question and run it

Research
What would you like to research?
“It provided multiple up-to-date articles and a table — exactly what I needed.”
Clinical pharmacist, NHS
“I like the fact that you see the reasoning, so you understand how it goes through all the information.”
Intensive care anesthesiologist, Italy
Your library in Causaly

The literature and the trial registries, searched in one question

40M+PubMed/MEDLINE abstractsPlus bioRxiv and medRxiv preprints
7M+Full-text articlesPubMed Central; Sage and Mary Ann Liebert on the Intelligence plan
550K+Clinical trialsRegistered and ongoing trials, searched with the literature
Drug labels and safetyDailyMed, OpenFDA, RxNorm
Society guidelines and the live webSearched when your question needs it
Your own documentsUpload your own documents for context or analysis
How it compares

Built for science, measured against general AI

Same research questions, three tools. Causaly finds more of the evidence that matters and cites nothing off-topic.

76%
Recall of relevant primary + supporting evidence

vs. 42% for ChatGPT and 20% for Claude

95.6%
Factuality on scientific outputs

with 100% groundedness in the evaluation

0%
Topically irrelevant citations

vs. 7% for ChatGPT and 30% for Claude

Explore the full benchmarking report
Your research stays yours

Your study ideas stay private

  • Never used to train models
  • Visible only to the people you share a workspace with
  • Encrypted in transit and at rest
  • ISO 27001 certified
Before you start
Research

Start your next study with the evidence in hand.

Get started on your own questions today.

Start your free trial

For research, literature review and evidence synthesis. Not for diagnosis or decisions about individual patients.