01 · Evidence Library

Thousands screened. The ones that matter, delivered.
Without weeks of winnowing.

We screen the published literature on your asset against the questions your project asks and place the papers that answer them, with the reason on every one. A library you read, inspect and overrule.

Where your writers start

Complex software gives your writers the work. We give them their time back.

Same five steps either way. The difference is how many of them your writers do, and where their time goes.

A tool you steer

Software you run yourself

the software medical writers tell us they struggle with
1
Set the protocolquestion, criteria, exclusion reasons
Your writers
2
Run and import the searchesexports from each database, de-duplicated by hand
Your writers
3
Screen every recordtwo reviewers, or one and a robot; conflicts adjudicated
Your writers
4
Tag, extract, appraiserecord by record
Your writers
5
Judge, select, exportthe decisions the document rests on
Your writers
5 of 5steps done by your writers.
Weeks of their time before the judging starts.
A library you receive

Posteros Evidence Library

with the work shown
1
Criteria written down before screeningthe questions the asset raises, fixed before screening
Done for you
2
Searches run for every questionrecorded, dated, de-duplicated
Done for you
3
Every record screened, with the reasonin or out, on the paper
Done for you
4
Every paper placed for its questionprimary, supporting or context, with the reason
Done for you
5
Judge, overrule, exportyour writers start here, with every decision on the record
Your writers
1 of 5steps done by your writers.
Their time goes into judgment, not winnowing.

Both are traceable. One records the decisions your writers made. The other records the decisions the system made, and hands your writers the overrule.

Step five

Step five has three screens.

One job each: read the library, inspect how it was built, select what you take forward.

Start here · 02 primary for clinical efficacy

Once-weekly treatment in adults with overweight or obesity: a 68-week randomised trial

Why this is in the library

The pivotal trial for the weight-reduction claim: 68-week primary endpoint against placebo, with the gastrointestinal event profile the safety section rests on. It does not report cardiovascular outcomes; those sit in a separate trial.

Reading questionWhich endpoint does your platform's efficacy claim rest on, and at what week?
Read original source ↗Inspect rationale →Mark readSave to reading list
Illustrative
✦ Read

Start by orienting to the domain.

Read opens on the papers that matter most for your objectives: the executive summary of everything we screened. From there, the full library, the clinical studies, the most cited. Every card says why the paper is in, in a sentence written for that paper, with the question it answers and a link to the source. Mark what you have read. Save what you will come back to. Search the library by meaning.

Collections start here · the complete library · clinical studies · most cited · my reading list.
On every card why it is in the library · the question it answers · its place for that question · the source.
✦ Inspect

Every system decision visible; you can overrule any of them.

Inspect is where a reviewer starts. Six cards answer the six questions a skeptical reviewer asks, with denominators: did you look everywhere, how much did you read, is it current, is it the right evidence, how hard were you on it, where did you use judgment. Click a card for the papers behind it. Click a paper for why it is in or out, and by how much.

On every paper the reason it is in or out · its place for each question · the margin ("ranked 4 of 178; the 13 kept scored 78–100") · authors, journal, year, source.
Your overrule on the paper itself, kept with the system's call, carried into the next edition.
Inspect six questions
Breadth
Did you look everywhere?
Depth
How much did you look at?
Currency
Is it current?
Quality
Right kind of evidence?
Selectivity
How hard were you on it?
Judgment
Where did you use judgment?
Select search by meaning, then export
cardiovascular outcomes with the once-weekly agents, no reviews · read as
Read as
− Narrative or scoping review+ Cardiovascular outcomes+ Clinical efficacy & safety
Export columns
Extracted findingsSource quotesResult directionLimitationsInclusion reasonReason for non-selectionScreening reasonsYour notesSnapshot datePlace for the question
Selecting for export does not change screening decisions →
Illustrative
✦ Select

Your writers make the final selections; we preserve the rationale.

Select is where the work leaves the library. Describe the evidence you need in plain words; the library turns it into filters you can see and change. Save sets. Export a dated Excel snapshot with the findings, the source quotes, each paper's place, and every reason, in and out. Your notes travel with it. Selecting changes no screening decision; the record stays whole.

Export Excel, dated, every reason included.
Saved sets kept in the library for the next person on the account.
Where it goes next the SCP Engine writes from this selection.
Four words

Every paper is in one of four states, and the state says why.

No internal labels on the surface. A paper is one of these, and the reason sits beside it.

CitedUsed in a document written from this library. The claim it supports is one click away.
In the libraryScreened in for at least one question, placed as primary, supporting or context, with the reason.
Considered, not includedPassed screening; not kept for the question. Always with a reason, often with the margin.
Screened outDid not pass the screen for any question. With the screening reason on the record.

Start from the library.
Everything else follows.

The SCP Engine writes from it. Evidence Delta re-runs it. Simulations read against it.

Talk to us All four products