Compare clinical trials point by point, with every difference cited
Two trials of the same treatment can reach different results because they enrolled different people, used a different comparator or measured the outcome at a different time. Upload the reports to one Search+ workspace and ask about one point of comparison at a time; each answer cites the passage from each report so you can see the difference for yourself.
Creating an account needs no payment.
Last updated October 2026
How to compare trial reports with Search+
- Upload every report you are comparing
Add each trial report or article, plus supplements where the detail lives, as PDFs, Word files or scans. Keep comparable trials in one workspace so a question covers all of them.
- List the comparison criteria once
In Workspace Context, write the points you will compare, for example "population, comparator, primary endpoint and timing, follow-up length, serious adverse events". Search+ applies that frame to every question.
- Ask one criterion at a time
Ask how the trials differ on a single criterion, then move to the next. Open the citations for each trial, because small wording differences, such as how an endpoint was defined, change what a result means.
Comparison questions to ask
How do the inclusion criteria differ between these trials, for example by age range or disease severity?
Which trials used placebo and which compared against an existing treatment?
Do the trials define the primary endpoint the same way and measure it at the same time point?
How long was follow-up in each trial, and how many participants were lost?
How do the reported rates of serious adverse events compare across the trials, and how did each define them?
Who funded each trial, and what conflicts of interest are declared?
Comparison points that matter for clinical trials
Different eligibility criteria produce different populations. A result in younger, healthier participants may not match one in an older or sicker group.
Placebo-controlled and active-controlled trials answer different questions, so their results are not directly interchangeable.
The same named endpoint can be defined or timed differently. Compare the definitions in the cited passages before comparing results.
Trials differ in how actively they looked for adverse events and how they graded severity, which affects reported rates.
Ask across every report in the workspace, or narrow to one, and the answer cites the passages it used from each report.
What to compare across trial reports
| Comparison point | Why it matters | A question to ask Search+ |
|---|---|---|
| Design | Randomized, blinded or open-label trials carry different risks of bias | Which of these trials were double-blind? |
| Population | Results apply to the people studied | How do baseline characteristics differ across the trials? |
| Intervention and dose | Different doses or schedules can change effects | What dose and schedule did each trial use? |
| Primary endpoint | Defines what success meant | How does each trial define its primary endpoint? |
| Follow-up length | Short trials can miss later effects or harms | How long was each trial's follow-up? |
| Analysis population | Intention-to-treat and per-protocol analyses can differ | Which analysis population did each trial report as primary? |
What is a trial comparison in Search+?
A trial comparison in Search+ is a chat answer that sets out how two or more trial reports in your workspace differ on the point you asked about, with citations to the passages in each report.
Questions about comparing clinical trials
Can Search+ tell me which trial is better?
Can it compare results from different trials?
Will it notice that two trials define an endpoint differently?
Is this medical advice?
How many reports can go in one workspace?
What if a comparison leaves a trial out?
See how the trials really differ
Start a workspace, upload the reports, and compare them one criterion at a time.
Start a workspace