Evidence tables, also called evidence matrices or synthesis tables, appear in EBP projects, literature reviews, PICOT papers and capstones. This guide shows which columns to use, how to extract study details accurately, how to rate evidence and how to turn the finished table into a written synthesis.
An evidence table organizes the studies you found into a grid, one row per source and one column per characteristic. It lets you, and your instructor, compare studies at a glance: who was studied, how, what was found and how strong the evidence is.
The table is a tool for analysis, not just a list. A well-built table shows patterns across studies, such as consistent findings, conflicting results or gaps in the evidence, which then become the basis of your written synthesis and recommendations.
Check your template first. Many nursing programs provide a required evidence table template, often based on the Johns Hopkins or a similar EBP model. If one is provided, use its columns exactly.
If no template is provided, these columns cover what most instructors expect:
| Column | What to record |
|---|---|
| Citation | Author(s), year, short title; full reference goes in the reference list |
| Purpose or research question | What the study set out to examine, in one sentence |
| Design | For example randomized controlled trial, cohort, qualitative descriptive, systematic review |
| Sample and setting | Number and type of participants, where and when |
| Intervention or variables | What was done or measured, and the comparison if any |
| Key findings | Main results relevant to your PICOT, with key numbers |
| Limitations | Main weaknesses noted by the authors or by you |
| Level and quality of evidence | Rating using your program's hierarchy |
| Relevance to your question | How the study applies to your population and setting |
For a narrow PICOT question you might add columns for outcome measures or follow-up period. Avoid adding so many columns that each cell becomes unreadably small.
| Weak cell entry | Strong cell entry |
|---|---|
| Findings: the intervention worked well | Falls fell from 4.1 to 2.6 per 1,000 patient days after hourly rounding (p < .05) |
| Sample: nurses and patients | 212 adult inpatients on two medical units, one urban teaching hospital |
| Limitations: some limitations | Single site; no control group; falls data from incident reports may undercount |
The figures in the strong column above are illustrative. In your own table, every number must come from the study you cite.
Most nursing programs ask you to rate both the level of evidence (based on design) and its quality (how well the study was done). Use the hierarchy your course specifies.
| Typical level | Design |
|---|---|
| Higher | Systematic reviews and meta-analyses of randomized controlled trials; individual RCTs |
| Middle | Quasi-experimental studies; cohort and case-control studies |
| Lower | Qualitative studies, descriptive studies, case reports |
| Non-research | Clinical practice guidelines, expert opinion, quality improvement reports |
Hierarchies number their levels differently. The Johns Hopkins model, for example, uses Levels I to V with quality ratings of A, B or C, while other models use up to seven levels. Name the model you used in your paper. For more detail, see our guide to nursing research critique.
Share your PICOT question, articles and template. A nursing writer can extract, rate and synthesize the evidence accurately.
Many students build a careful table and then write a summary that describes each study one by one. Instructors want synthesis: what the studies show together.
Synthesis sentence example: "Four of the six studies, including two randomized trials, found that structured hourly rounding reduced falls on adult medical units, while the two studies that found no effect had short follow-up periods and low adherence to the rounding protocol."
Follow your assignment. Course-level EBP papers often require five to ten studies; capstone and DNP projects may require more. Recency requirements are common, often within the last five years, with exceptions for landmark studies or guidelines if your instructor allows them. Quality matters more than quantity: a table of eight relevant, well-rated studies is stronger than fifteen loosely related ones.
Most students build evidence tables in Word, but a spreadsheet can make the extraction stage faster. Create one column per characteristic, fill in a row as you read each article, then sort by design or theme to see patterns before copying the final version into an APA-formatted table in your paper.
If your assignment asks you to describe your search, note the databases, search terms, limits such as publication years, and how many articles you screened and kept. Our nursing literature review guide covers search strategy in more detail.
No. An annotated bibliography gives a paragraph per source. An evidence table compares sources side by side in a grid, which makes patterns and differences easier to see. Some assignments ask for both.
Usually yes, if your assignment allows non-research evidence. Label them clearly and rate them with your model's non-research criteria.
Follow your instructions. Large tables are often placed in an appendix and referred to in the text, while some assignments want the table within the body.
Short and specific: a phrase or one or two brief sentences. The table should be scannable, with detailed discussion saved for the written synthesis.
The one your program specifies. Johns Hopkins and Melnyk and Fineout-Overholt hierarchies are common in nursing. Name the model in your paper.
A strong evidence table is accurate, concise and honest about limitations, and it leads naturally to a synthesis that answers your clinical question. Build it carefully from full-text articles, rate every study with a named model, and let the patterns in the table drive your recommendations.