PRISMA-style evidence table template
An evidence table is where a review stops being a reading list and starts being a defensible argument. This template gives you the columns, an illustrative row, a source reliability rubric, and the screening flow — free, no account needed. Keep the filled table in Evidence Pilot when you want the decisions and the audit trail together.
The thirteen columns
- Study ID — a short stable label you can cite in the text (for example, the first author and year).
- Citation — the full reference, or a link to the record in your reference library.
- Study design — the design as the authors describe it: trial, cohort, case–control, survey, qualitative, modeling, and so on.
- Population / sample — who was studied, how many, and any inclusion limits that matter to your question.
- Intervention / exposure — what was done to or measured in the study.
- Comparator — what it was compared against, including “none” when there was no control.
- Outcomes measured — the outcomes as measured, not the outcomes you wish were measured.
- Effect size / result — the direction and magnitude, with uncertainty when the source reports it.
- Risk of bias — your judgment of the main threat to the result, with the reason.
- Relevance to the review question — direct, partial, or indirect, and why.
- Decision — include or exclude, recorded explicitly.
- Reason for decision — the criterion that drove the decision, not a restatement of it.
- Notes — reviewer comments, follow-ups, and anything the next reader needs.
The table
The row below is illustrative, not a real study. Replace it with your first record.
| Study ID | Citation | Study design | Population / sample | Intervention / exposure | Comparator | Outcomes measured | Effect size / result | Risk of bias | Relevance | Decision | Reason for decision | Notes |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| EX-001 | Example et al. (2026). Illustrative record — replace this row. | Randomized trial (illustrative) | n=120 adults (illustrative) | Example intervention | Usual care | Primary outcome at 12 weeks | Direction and confidence interval as reported by the source | Moderate: allocation described, outcome assessor not blinded | Direct | Include | Meets population and design criteria | Check the full text before relying on the result |
Source reliability rubric
Score each source on five criteria from 1 (weak) to 5 (strong). The rubric is a prompt for judgment, not a substitute for it; record the reason for any score below 3.
| Criterion | What to ask | Score (1–5) |
|---|---|---|
| Authority | Is the author, institution, or publisher a recognized source for this kind of claim? | |
| Currency | Is the source recent enough for the claim it supports, and is there a newer correction or retraction? | |
| Methodology | Is the method described clearly enough to judge, and is it appropriate to the question? | |
| Independence | Is the source free of an obvious stake in the conclusion (funding, advocacy, self-report)? | |
| Corroboration | Is the finding supported by at least one independent source? |
Total (out of 25): 21–25 strong · 16–20 usable with stated caveats · 11–15 weak, corroborate before relying · 10 or below, do not rely on the source alone. These bands are a starting convention; your methods reviewer sets the final threshold.
PRISMA-style screening flow
Record a count at every stage. Counts that do not reconcile are the most common reason a review cannot be reproduced.
- Identification — records identified from databases and registers.
- Duplicates — records removed before screening.
- Screening — records screened at title and abstract; records excluded, with reasons.
- Eligibility — full-text reports assessed; reports excluded, with reasons.
- Included — studies included in the review, and reports of those studies.
From template to source trail
A spreadsheet holds the table. It does not hold the claim each row supports, the screening decision history, or who reviewed what. Evidence Pilot keeps the evidence matrix, screening queue, confidence bands, and audit receipts together on the live project — and exports a PRISMA-style flow plus references when you need to hand the review to someone else. If you are choosing between tools, read organize vs. reason.
Common questions
Is this evidence table template PRISMA-certified?
No. PRISMA is a reporting guideline, not a certification. This template follows the PRISMA-style screening flow and evidence-table structure so a review is easier to report and audit. You still apply the guideline that fits your field and review type, and your methods reviewer makes the final call.
Can I use these templates without an Evidence Pilot account?
Yes. The evidence table, source reliability rubric, and screening flow checklist are free to download and use. Evidence Pilot is where you keep the table, the decisions, and the audit trail together on a live project.
How is an evidence table different from a bibliography?
A bibliography lists sources. An evidence table records what each source found, how it was designed, who it studied, how strong the result was, and whether it was included or excluded and why. The table is the reasoning layer a bibliography cannot carry.