What Columns Should an Evidence Table for a Systematic Review Include? (Template)

TL;DR

There is no universal perfect template, but most defensible evidence tables share the same core column groups.

This post gives:

a practical starter template
required vs optional column guidance
formatting rules that keep tables synthesis-ready

First principle: design columns for synthesis, not extraction convenience

Before defining columns, confirm:

target decisions the review must support
planned synthesis type (quantitative, qualitative, mixed)
key subgroup/timepoint requirements

A column exists only if it supports comparison, synthesis, or decision defense.

For outcome strategy, see How to Choose Outcomes for an Evidence Table: Quantitative vs Qualitative Reviews.

Core column groups (starter template)

1) Study identification

author/year
citation ID/PMID
country/setting
funding/conflict declaration

2) Study design and methods

design type (RCT, cohort, etc.)
arms/comparators
sample size by arm
follow-up duration

3) Population

eligibility criteria summary
baseline age/sex and key severity marker
relevant subgroup flags

4) Intervention and comparator details

dose or device version
duration/intensity
co-interventions

5) Outcomes (definition layer)

exact outcome definition
measurement instrument
analysis set (if relevant)
planned timepoint window

6) Results (numeric layer)

effect estimate
confidence interval in separate columns
event counts/denominator where needed
optional p-value

7) Quality/risk of bias

tool used
overall/domain judgment
key caveat note

8) Traceability and notes

source location/snippet reference
extraction caveat
adjudication note if changed

Required vs optional fields

Usually required

study design
comparator clarity
outcome definition
timepoint
effect estimate
confidence interval or equivalent uncertainty metric

Context-dependent

funding/conflict
subgroup depth
detailed protocol deviation notes
implementation context details

Start lean but never omit fields required for interpretation.

Formatting rules that save projects

one variable per column
one unit convention per numeric column
separate baseline and follow-up values
separate estimate and CI columns
explicit missingness values (not blank cells)

These rules improve QA, filtering, and reproducibility.

Minimal quantitative template example

Group	Columns
Study ID	Author/year, PMID, country
Design	Design type, arms, N, follow-up
Population	Eligibility summary, baseline age, subgroup flag
Intervention	Intervention detail, comparator detail
Outcome definition	Endpoint, instrument, timepoint
Results	Effect estimate, CI lower, CI upper, events/denominator
Quality	Tool, rating, key caveat
Traceability	Source pointer, extraction note

Use this as a starting point, then tailor to your synthesis plan.

Common column design mistakes

Over-collecting non-usable variables

Creates extraction burden without improving synthesis.

Composite text cells

Breaks analysis and complicates QA.

Missing timepoint policy

Leads to hidden non-comparability.

Treating quality as a single score only

Hides domain-level concerns needed for interpretation.

Final thought

A good evidence table is not about having many columns. It is about having the right columns, defined clearly, for the analysis you actually plan to run.