Skip to content
ClinicalEvidence

AI for doctors who need evidence they can open and check

ClinicalEvidence is AI for doctors and clinical pharmacists that answers clinical questions only from registered trials and published studies. Every sentence carries an NCT number or PMID you can open, and when the evidence is not there, the answer says so plainly.

Find evidence
Researcher typing beside printed study pages

Why a general chatbot is risky for clinical evidence

General purpose chatbots are trained to produce fluent text, not to quote sources. Ask one what the trials show for a drug in a specific population and you often get a confident paragraph with references that look real: a plausible journal, a plausible year, a plausible author list. Some of those studies do not exist. Others exist but say something different from the sentence that cites them.

For a physician or a clinical pharmacist this is not a style problem. A reference you cannot verify is a reference you cannot use in a case discussion, a formulary review, a reply to a colleague or a teaching session. Checking every claim by hand takes longer than doing the search yourself, which removes the reason to ask in the first place.

  • References that point to studies that were never published
  • Real studies cited for findings they did not report
  • Answers written from model memory when no evidence exists
  • No record of which sources were actually searched

Read more on why AI tools invent citations.

Cited answers built only from trials and studies

ClinicalEvidence works the other way around. It first searches its own index of registered trials and published studies (ClinicalTrials.gov, PubMed, PubMed Central open access articles, the EU Clinical Trials Register and WHO ICTRP where licences allow), refreshed every day. The language model then writes the answer only from the passages it retrieved, never from memory.

Before you see anything, each citation goes through automatic verification. The system checks that the NCT number or PMID exists in the index and that the cited passage actually supports the sentence. Sentences that fail are removed. If nothing survives, you get the matching studies and a plain statement that no verified answer exists.

  1. Answers grounded in retrieved passages from trial records and abstracts

  2. Every sentence linked to its source record

  3. No verified source, no answer

  4. A "Sources searched" panel that lists only the sources used for your question

See exactly how cited answers are verified.

What you see after you ask

You type the question the way you would ask a colleague, for example whether SGLT2 inhibitors reduce hospitalisation in heart failure with preserved ejection fraction, or which phase 3 trials test a new anticoagulant in patients with atrial fibrillation and kidney disease. You can narrow the search to registered trials, publications, a study phase or completed studies with results.

Next to the written answer you get a trial table with phase, population, interventions, primary and secondary endpoints, results posted, sponsor, status, enrolment, dates and locations. You can sort and filter it by condition, intervention, phase, status, study type, date range, sponsor type, country and results available.

  • Short answer with an NCT number or PMID on every sentence
  • Level of evidence per cited source in evidence summaries (randomized trial, meta-analysis, cohort, case series)
  • Side by side comparison of two to four interventions
  • Email alerts when new trials or publications match a saved question
  • Export to CSV, RIS and BibTeX on every plan

Built for physicians and clinical pharmacists in professional practice

ClinicalEvidence is a professional tool for people who read evidence for a living: hospital physicians, specialists in group practice, clinical pharmacists working on formularies and drug information, residents preparing journal clubs and clinicians who teach.

It is not for patients and it is not medical advice. It does not see or process patient data, it does not suggest a diagnosis or a treatment for a person, and it does not replace clinical judgment. It shows you what the registered trials and published studies report, with the source attached, so that you can read, weigh and decide. Read our medical disclaimer for the full scope.

  • Checking what trials show before a treatment discussion
  • Preparing drug information answers and formulary reviews
  • Building journal club and teaching material with real references
  • Following new trials in your specialty with alerts

Plans for clinicians and clinical teams

Every plan includes cited answers, citation verification, the trial table and intervention comparisons. Save 50% with yearly billing.

  • Clinician 49 USD per month, or 24 USD per month billed yearly (288 USD), 200 cited answers per month and 1 seat
  • Research Team 199 USD per month, or 99 USD per month billed yearly (1 188 USD), 800 pooled answers and 3 seats
  • Organization 499 USD per month, or 249 USD per month billed yearly (2 988 USD), 2 000 pooled answers and 10 seats
  • Enterprise 1 490 USD per month, or 745 USD per month billed yearly (8 940 USD), 6 000 pooled answers and up to 50 seats

Compare every limit on the pricing page.

Questions doctors ask about ClinicalEvidence

How is ClinicalEvidence different from a general chatbot?

A general chatbot writes from what it learned during training and can produce references that do not exist. ClinicalEvidence answers only from passages retrieved from its index of trials and studies, links every sentence to an NCT number or PMID and removes any sentence whose citation fails verification.

What happens when there is no evidence for my question?

You get a plain statement that no verified answer exists, together with the closest matching studies. ClinicalEvidence never fills the gap from model memory.

Can I enter patient details?

No patient data is needed and you should not enter any. Ask about a population, an intervention and an outcome, the way a study would define them.

Does ClinicalEvidence make treatment recommendations?

No. It reports what registered trials and published studies show, with sources. Clinical decisions stay with you and your team.

Can I try it before subscribing?

Yes. The homepage demo answers one question with real citations and the first three matching studies. After that, ClinicalEvidence is a paid subscription.

Ask your clinical question and open every source

One question, a cited answer in seconds, and a trial table you can check line by line.