ProgniaPrognia
Cited Answers

Trust Is a Link, Not a Promise

Osler answers clinical questions with numbered inline citations drawn from eleven source families — guidelines, PubMed, Cochrane, trials and more — each one click from the primary source.

Osler
Cited clinical AI · 11 source families
Is apixaban preferred over warfarin in elderly patients with AF?
Searching 11 source families…
3 sources · guidelines, literature, trials

The problem

A Confident Answer Is Not Evidence

Generic chatbots answer medical questions fluently and source-free. In medicine, an answer you cannot verify is an answer you cannot use.

  • 1

    No sources, no safety

    A generic chatbot delivers medical claims with total confidence and zero references. You are left guessing which parts are real.

  • 2

    Verification costs more than the answer

    Checking an unsourced claim means searching PubMed and the guidelines yourself — the exact work the AI was supposed to save.

  • 3

    Yesterday’s medicine, presented as today’s

    Without a cited society and year, you cannot see whether an answer reflects current guidance or a version that has since been replaced.

Inline citations

Every Claim Carries Its Sources

Osler answers with numbered citation chips inline — [1], [2], [3] — placed exactly where each claim is made. Hover a chip for a source card; click it to open the full sources sheet, scrolled to that source.

  • Numbered citation chips inside the answer text
  • Hover cards show source type, title and sub-label
  • A full sources sheet behind every answer
  • Citations rendered in every single answer

Osler’s answer

DOAC over VKA is a Class I recommendation— in ARISTOTLE, apixaban cut major bleeding by 31% versus warfarin.

Sources

3 cited

Eleven source families

Eleven Kinds of Evidence, One Answer

Guidelines, PubMed, Europe PMC, Cochrane, FDA drug labels, DailyMed, EMA, ClinicalTrials.gov, drug interactions, RxNorm and WHO ICD-11 — each family color-coded consistently, so you can read an answer’s evidence at a glance.

  • Society guidelines cited with their year
  • Literature from PubMed, Europe PMC and Cochrane
  • Drug data from FDA labels, DailyMed and EMA
  • Trials from ClinicalTrials.gov with phase and status

Guidelines

ESC 2023 · Acute coronary syndromes

Primary sources

One Click to the Original

Every source deep-links to the primary record — PubMed by PMID, DOI, the ClinicalTrials.gov study page, FDA and DailyMed labels. You verify at the source, not at a summary of a summary.

  • PubMed articles opened by PMID
  • DOIs resolve to the original paper
  • Trial registry pages linked directly
  • Drug labels opened at FDA and DailyMed
PubMed

Apixaban versus Warfarin in Patients with Atrial Fibrillation

Granger et al. · N Engl J Med 2011

pubmed.ncbi.nlm.nih.gov/21870978

PubMed abstract · PMID 21870978

The original record — not a paraphrase

A real conversation

History That Stays With You

Osler is a conversation, not a search box. Recent context carries across turns, and every conversation is saved and searchable — so a cited answer from last month is still one search away.

  • Follow-up questions keep their context
  • Every conversation persisted automatically
  • Search across your full conversation history
  • Pick any thread back up where you left it
Search your conversations

All conversations

Apixaban vs warfarin in elderly AF

12 May · 4 sources

Apixaban dosing in CKD stage 4

3 Jun · 6 sources

Perioperative bridging for mechanical valves

21 Jun · 5 sources

Statin intolerance — rechallenge or switch?

28 Jun · 3 sources

Eleven Source Families, One Answer

Every family is color-coded consistently, so you can read the evidence behind an answer at a glance.

Clinical guidelines

ESC 2023 · Acute coronary syndromes

Society and year on every citation

Generic Chatbot. Cited Clinical AI.

The difference between an answer you can repeat and an answer you can verify.

Generic AI chat

  • Fluent claims with no references
  • “Trust me” as the evidence model
  • You re-search everything to verify it
  • Trained knowledge with an unclear cutoff
  • Chats vanish when the tab closes
  • One-shot answers with no memory

With Osler

  • Numbered citations inline, on the claim itself
  • Eleven source families, color-coded
  • One click to PubMed, the DOI or the trial registry
  • Guidelines cited with society and year
  • History saved and searchable
  • Context carried across the conversation

From Question To Primary Source

Four steps between asking and verifying — none of them a leap of faith.

  1. 1

    Ask

    Type a clinical question the way you would ask a colleague — no query syntax, no forms.

  2. 2

    Osler retrieves

    Relevant sources are pulled from eleven families: guidelines, literature, drug data, trials.

  3. 3

    Read the cited answer

    Claims arrive with numbered chips; hover any chip to see exactly what stands behind it.

  4. 4

    Verify at the source

    Click through to PubMed, the DOI or the registry — the primary record, not a paraphrase.

Connected to the Prognia ecosystem

The Path of a Cited Answer

Every answer follows the same route — retrieval first, citation always, primary source one click away.

Your clinical question
Eleven source families
Cited answer
Primary source

Questions about Cited Answers

Ask Anything. Verify Everything.

Clinical questions answered with inline citations from eleven source families — start free with 3 cited chats a day.

Free plan included · No credit card required