FREE GUIDE
Title and abstract screening guide
A step-by-step workflow for new reviewers. Practice decisions on five example records, then run the same process on a free EvidenceFlow project.
- 1. Lock criteria before you open EndNote. Write inclusion and exclusion as yes/no tests against PICO. If two reviewers would disagree on a borderline abstract, the criterion is still too vague.
- 2. De-duplicate, then screen titles. Remove duplicate records first so PRISMA identification numbers stay honest. Title screening is only for obvious excludes (wrong species, conference tweet, clearly off-topic). Keep anything uncertain.
- 3. Screen abstracts independently. Two reviewers, blinded to each other if your protocol requires it. Allowed decisions: include, exclude (with a reason code), or unsure. Unsure always goes to full text — never to exclude.
- 4. Resolve conflicts on a schedule. Do not re-screen the whole set together. Meet on disagreements only. Record who decided and why. That log is what editors ask for when dual review is claimed.
- 5. Retrieve full text and repeat. Track reports not retrieved separately from reports excluded after reading. Full-text excludes need a PRISMA reason (wrong population, intervention, comparator, outcome, or design).
- 6. Count as you go. Identified, after duplicates, screened, excluded, sought, not retrieved, assessed, excluded with reasons, included. Those eight numbers are the flow diagram. Do not reconstruct them from memory at write-up.
Practice worksheet (not saved on a server)
SGLT2 inhibitors and heart failure hospitalization in adults with T2D: a multicenter RCT
Mouse model of empagliflozin and cardiac fibrosis
Narrative review of glucose-lowering drugs in primary care
Dapagliflozin versus placebo for HbA1c: 26-week double-blind trial
Conference abstract: SGLT2i in 12 patients (no methods)
Methods write-up: Title/abstract screening methods guide
What good screening looks like
Title/abstract screening is a filter, not a verdict on study quality. You are asking only whether the paper could meet eligibility. Risk of bias, sample size, and “this journal is weak” are not exclusion reasons at this stage.
Dual independent screening plus a documented conflict process is what most protocols promise. The worksheet below is a rehearsal; EvidenceFlow records the same include/exclude trail against imported RIS, BibTeX, or PubMed sets.
FAQ
Frequently asked questions
Should I exclude based on title alone?+
Only when the title makes eligibility impossible (wrong species, clearly the wrong intervention). If the title is merely vague, keep it for abstract or full text.
What if the abstract is missing?+
Treat it as unsure and retrieve the record. Missing abstracts are common in older conference items and should not be silently dropped.
Do both reviewers need to agree on every include?+
Includes that either reviewer flags usually proceed to full text. Conflicts on exclude vs include should be discussed; do not default to the more senior reviewer's gut feel without a recorded reason.
Is AI allowed to make the decision?+
A relevance score can reorder the queue. The include/exclude call stays with the human reviewers. EvidenceFlow's core screening works without paid AI.
Import a RIS or PubMed set and screen with dual reviewers on the free core workflow. AI ranking is optional and paid; the decision log is not.
Start free project