From search results to
systematic review — faster.

LitRev combines AI screening, PICOS-based reasoning, and structured data extraction — so medical researchers can complete rigorous reviews without the manual burden.

Start your first review See how it works

Free for 1 project · Used by 4,000+ researchers

app.litrev.io/reviews/cv-rct-meta-analysis/screening

LitRev

Projects

CV RCT meta-analysis

Diabetes scoping review

Current review

Title/abstract screening

Full-text review

Data extraction

Export

Title / abstract screening

CV RCT meta-analysis · 4,821 citations

AI screening active

312

Included

4,461

Excluded

Pending

99%

Complete

✓

Statin therapy in adults with T2DM: intensive vs standard lipid-lowering — multicentre RCT

Chen et al. · The Lancet · 2023

AI · Include (96%) — Meets all PICOS: adult T2DM population, statin intervention, cardiovascular outcomes, RCT design.

✗

PCSK9 inhibitors in paediatric familial hypercholesterolaemia: a randomised trial

Müller et al. · JACC · 2023

AI · Exclude (94%) — Population criterion not met. Paediatric subjects aged 10–17; protocol requires adults ≥18.

—

Long-term cardiovascular outcomes after intensive glycaemic control: ACCORD 10-year follow-up

ACCORD Study Group · NEJM · 2023

Trusted at

University of OxfordJohns Hopkins MedicineCleveland ClinicStanford UniversityKing's College of LondonUniversity of Edinburgh

The problem

Systematic reviews are
slow by design. Not by necessity.

The traditional workflow hasn't changed in decades. Researchers spend months on tasks that don't require human judgement.

Manual screening at scale is unsustainable

Screening thousands of citations by hand takes weeks, introduces fatigue-based errors, and requires two independent reviewers for every record.

“We screened 11,000 abstracts over three months. Two of us. It was exhausting.”

Extraction is inconsistent across reviewers

Without structured templates and validation, different reviewers extract different data from the same paper — leading to reconciliation overhead and unreliable synthesis.

“We discovered two reviewers were interpreting ‘sample size’ differently. Halfway through extraction.”

No audit trail for peer review

Journal reviewers and PROSPERO registrations require a complete record of every inclusion decision. Spreadsheets can't provide that — and cobbled-together workflows leave gaps.

“The reviewers asked for our exclusion rationale per record. We had nothing logged.”

How it works

Five steps. One platform.

From search strategy to publication-ready output, every stage of the review workflow is covered.

Import

Connect to PubMed, Embase, Cochrane, and 5 other databases, or upload RIS / CSV files directly.

Screen

AI screens every title and abstract against your PICOS criteria. Review AI decisions with full reasoning.

Full-text

Upload PDFs for full-text review. AI flags eligibility concerns by section and page reference.

Extract

Define extraction fields. AI pre-fills values from full text. Reviewers confirm or override each field.

Export

Generate PRISMA flow diagrams, Word reports, CSV data tables, and complete audit logs.

Explainable AI

AI that shows its reasoning.

Every AI decision in LitRev is explained using your study's PICOS criteria — never a black box.

✦ AI screening

Title & abstract screening with PICOS logic

Each citation is evaluated against every element of your PICOS framework. Confidence scores and criterion-level explanations are provided for every decision.

Sample reasoning

Include (91%) — RCT recruiting adults with T2DM (P ✓). Compares metformin vs lifestyle intervention (I/C ✓). Primary outcome: HbA1c at 12 months (O ✓). Study design: parallel RCT (S ✓).

P — metI — metC — metO — metS — met

✦ AI extraction

Structured data extraction from full text

Upload a PDF and AI locates each extraction field across methods, results, and appendix sections — citing the exact sentence and page number for every value.

Sample extraction

Sample size: 342 (p.3, Methods) · Follow-up: 52 weeks (p.4, Table 1) · HbA1c reduction: −1.4% (95% CI −1.8 to −1.0) (p.7, Results) · Risk of bias: Low — allocation concealed (p.12, Appendix)

✦ AI synthesis

Evidence summaries grounded in included studies

After extraction, AI generates narrative summaries and evidence tables — citing specific studies and noting heterogeneity, gaps, and implications for practice.

Sample synthesis

Across 14 included RCTs (n=4,821), intensive glycaemic control was associated with reduced HbA1c (pooled WMD: −1.2%; I²=34%). Evidence for cardiovascular benefit remains limited — only 3 trials reported MACE as a primary outcome.

Comparison

How LitRev compares

Built for teams that need more than basic screening tools.

Feature	LitRev	Covidence	Rayyan
AI title/abstract screening	✓ With PICOS reasoning	—	Limited (no explanation)
Explainable AI decisions	✓ Criterion-level	—	—
AI data extraction from PDF	✓ With source citations	Basic (no AI)	—
AI narrative synthesis	✓	—	—
Multi-database search builder	✓ 8 databases	Import only	Import only
Full audit log per decision	✓	✓	Partial
PRISMA flow diagram export	✓ Auto-generated	✓	Manual
Team collaboration	✓	✓	✓
Free tier	✓ 1 project	Trial only	✓ Limited
Starting price	$49 / month	$99 / month	Free / $10 month

Pricing

Simple, transparent pricing

Start free. Upgrade when your research demands it.

Free

forever

✓1 project

✓Up to 200 citations

✓Manual title/abstract screening

✓Full-text review

—AI screening

—AI data extraction

—AI synthesis

—Team collaboration

Get started free

Pro

Your next systematic review starts here.

Join 4,000+ researchers who've moved past spreadsheets, manual screening, and fragmented workflows.

Start your first review Request a demo

Free for 1 project · No credit card required

From search results tosystematic review — faster.