OpenEvidence is an AI platform built to answer clinical questions grounded in scientific literature, with citations. It shortens the path to studies and guidelines, organizes evidence, and helps you compare alternatives clearly. This guide focuses on practical use, study and shift routines, ready-made prompts, and safety checks.

What changes in practice

The platform's biggest value shows up in three concrete gains: speed to source, organized comparison, and traceable reasoning. Instead of ten open browser tabs, you start from a structured answer and go straight to the studies backing the argument.

Where it shines: comparing treatment approaches and options · checking a guideline and its exceptions · understanding outcomes, benefit, and risk · preparing a case discussion or round · studying for evidence understanding, not memorization.

What's poor use: a vague question with no clinical context · accepting a summary without opening the primary references · using it as an automatic decision-maker · ignoring applicability to your specific patient and local context.

The workflow that delivers the most value

1. Write the clinical scenario in one sentence
2. Ask a structured question
3. Request an answer format that helps you decide
4. Open the primary references and validate applicability
5. Turn the evidence into a clear plan for your context

How to ask better questions: ready-to-copy templates

A good question has context and an outcome. The goal isn't "what's best" — it's "under which conditions does A beat B, and at what risk cost."

Short PICO template: Patient/context: [age/condition/comorbidities/setting]. Intervention: [A]. Comparison: [B or standard]. Outcome: [mortality, admission, complication, control, quality of life]. Request: bullet-point answer + 5 primary references + evidence level.

A vs. B comparison prompt: "Compare A vs. B for [condition] in [population]. Answer in 6 points: who benefits most, key outcomes, risk profile, guideline stance, exceptions, and practical bottom line."