Systematic Review Abstract: A PRISMA-Aligned Worked Example

A systematic review abstract has a different job than an abstract for a single primary study. Because a systematic review synthesizes many studies rather than reporting one, its abstract has to compress a search strategy, an eligibility screen, a risk-of-bias assessment, and a synthesis of results into a few hundred words — and readers, editors, and indexing databases increasingly expect it to do so in a fixed, labeled structure. That structure is not arbitrary: it comes from the PRISMA 2020 statement, which publishes a companion checklist specifically for systematic review abstracts. This guide walks through that checklist section by section and shows what it looks like applied to a full worked example.

Why systematic review abstracts follow PRISMA, specifically

Page MJ, McKenzie JE, Bossuyt PM, et al., “The PRISMA 2020 statement: an updated guideline for reporting systematic reviews,” BMJ 2021;372:n71, is the current reporting standard for systematic reviews and meta-analyses, and is required or recommended by most major biomedical and many social-science journals. Alongside the main 27-item checklist and flow diagram, PRISMA 2020 publishes a separate, shorter companion checklist — PRISMA 2020 for Abstracts — built specifically for the abstract, because an abstract’s word-count constraints make the full reporting checklist impractical to apply directly. The ICMJE’s own manuscript-preparation recommendations reinforce the same expectation from the publishing side: structured abstracts are required for original research, systematic reviews, and meta-analyses, though ICMJE leaves the exact subheadings and word limit to each journal (ICMJE Recommendations).

The PRISMA 2020 for Abstracts checklist

The checklist has 12 items, organized under the same four headings the main PRISMA checklist uses. A systematic review abstract that addresses all 12 gives a reader enough information to judge the review’s scope and reliability without opening the full paper:

1. Title — identify the report as a systematic review (or meta-analysis, if applicable).
2. Background: Objectives — state the main objective(s) or question(s) the review addresses.
3. Methods: Eligibility criteria — specify the inclusion and exclusion criteria for the review.
4. Methods: Information sources — specify the databases, registers, or other sources searched, and when each was last searched.
5. Methods: Risk of bias — specify the methods used to assess risk of bias in the included studies.
6. Methods: Synthesis of results — specify the methods used to present and synthesize results.
7. Results: Included studies — give the total number of included studies and participants, and summarize relevant study characteristics.
8. Results: Synthesis of results — present results for the main outcomes, ideally with the number of studies/participants contributing to each; report summary estimates and confidence intervals if meta-analysis was performed, and the direction of effect if comparing groups.
9. Discussion: Limitations of evidence — summarize the limitations of the included evidence (e.g., risk of bias, inconsistency, imprecision).
10. Discussion: Interpretation — give a general interpretation of the results and their implications.
11. Other: Funding — state the primary source of funding for the review itself.
12. Other: Registration — give the register name and registration number (e.g., a PROSPERO registration number).

Journals rarely reproduce these 12 items as 12 separate subheadings — most collapse them into a shorter, named set of structured sections. The worked example below uses one of the most common versions of that collapsed structure.

Section-by-section: what each part of the abstract is doing

Background

One or two sentences establishing the clinical, scientific, or policy problem and the specific gap the review fills. This is not a literature review in miniature — it states why the synthesis was needed, not what every prior study found.

Objectives

A single explicit sentence stating the review’s question, ideally in a PICO-style form (population, intervention/exposure, comparison, outcome) so a reader can immediately judge whether the review is relevant to their own question.

Data Sources

The databases and registers searched (e.g., MEDLINE, Embase, Cochrane CENTRAL, trial registries) and the search end date. This maps to PRISMA abstract item 4 (information sources) and matters because a reader assessing currency needs to know how recent the evidence base is.

Study Eligibility Criteria

The inclusion and exclusion rules applied during screening — study design, population, exposure/intervention, comparator, outcome, language, and date restrictions, as applicable. Vague criteria here are one of the most common weaknesses in submitted systematic review abstracts, because they make it hard for a reader to judge whether the review’s scope matches their own.

Methods (appraisal and synthesis)

A brief statement of how risk of bias was assessed (e.g., the Cochrane Risk of Bias tool, ROBINS-I, or another named instrument) and how results were synthesized — meta-analysis with a stated model, or narrative/qualitative synthesis if pooling wasn’t appropriate.

Results

The number of studies and total participants included, key study characteristics, and the synthesized findings for the primary outcome(s). If a meta-analysis was performed, this is where a summary effect estimate and confidence interval belong.

Limitations

A candid, brief statement of what weakens confidence in the findings — risk of bias across included studies, heterogeneity/inconsistency, imprecision from small sample sizes, or gaps in the search. PRISMA explicitly asks for limitations of the evidence, not limitations of the review’s own conduct, though many abstracts cover both in one or two sentences.

Conclusions

What the synthesized evidence supports, stated at a level of certainty that matches the Limitations section directly above it — a review with serious risk-of-bias and inconsistency concerns should not conclude with unqualified confidence. This section commonly closes with the review’s practical or research implications and, per PRISMA abstract items 11 and 12, the funding source and registration number (often a PROSPERO ID).

Worked example (illustrative composite)

This is an illustrative composite example written for teaching purposes only. It does not describe a real systematic review, cites no real studies, and reports no real data. It is built solely to show what a PRISMA-aligned structure looks like filled in, the same way a filled-out form illustrates a form’s fields. Any resemblance to an actual published review is coincidental.

Effectiveness of Telehealth-Delivered Cognitive Behavioral Therapy for Adult Insomnia: A Systematic Review and Meta-Analysis

Background: Cognitive behavioral therapy for insomnia (CBT-I) is recommended as first-line treatment, but in-person delivery is limited by clinician availability. Whether telehealth-delivered CBT-I achieves comparable effectiveness to in-person delivery remains unclear.

Objectives: To synthesize evidence on the effectiveness of telehealth-delivered CBT-I compared with in-person CBT-I or waitlist/usual care for adults with chronic insomnia, measured by change in Insomnia Severity Index (ISI) score.

Data Sources: MEDLINE, Embase, PsycINFO, and Cochrane CENTRAL were searched from database inception to March 2026, supplemented by a search of ClinicalTrials.gov and manual reference screening of included studies.

Study Eligibility Criteria: Randomized controlled trials enrolling adults (18+) with chronic insomnia, comparing telehealth-delivered CBT-I (video or telephone) with in-person CBT-I or a waitlist/usual-care control, reporting ISI or a validated equivalent, published in English.

Methods: Two reviewers independently screened records and assessed risk of bias using the Cochrane Risk of Bias 2 tool. Random-effects meta-analysis was used to pool standardized mean differences (SMD) in ISI change scores; heterogeneity was assessed with I².

Results: Fourteen trials (n=2,183) met inclusion criteria. Telehealth-delivered CBT-I produced significantly greater ISI reduction than waitlist/usual care (SMD −1.12, 95% CI −1.34 to −0.90; 11 trials) and was not significantly different from in-person CBT-I (SMD 0.08, 95% CI −0.11 to 0.27; 6 trials). Heterogeneity was moderate (I²=48%).

Limitations: Most included trials had unclear or high risk of bias for blinding of outcome assessment, given the nature of the intervention; follow-up beyond 6 months was reported in only 4 trials, limiting conclusions about durability.

Conclusions: Telehealth-delivered CBT-I appears comparably effective to in-person delivery and superior to waitlist/usual care for adult chronic insomnia, though long-term durability evidence remains limited. This review received no external funding and was registered with PROSPERO (illustrative registration number: CRD00000000).

Word count and format vary by journal — don’t assume one standard

ICMJE does not set a single word-count ceiling for structured abstracts; it leaves the specific format and length to each journal’s instructions for authors. In practice, most journals cap systematic review abstracts somewhere in the 250–350 word range, and the exact subheadings above are one common version, not a universal one — some journals use PRISMA’s own four-part grouping (Background/Objectives, Methods, Results, Discussion) instead of the eleven-part expanded version shown here. Always check the target journal’s own author instructions rather than assuming a fixed format; PRISMA’s abstract checklist tells you which pieces of information need to appear somewhere in the abstract, not what to name each subheading.

Common mistakes

• Omitting the registration number. PRISMA abstract item 12 asks for it explicitly; leaving it out is one of the most frequently flagged gaps in editorial review.
• Vague eligibility criteria. “Relevant studies were included” tells a reader nothing checkable; state the actual population, comparator, and outcome restrictions.
• Reporting a single number without dispersion. An effect estimate without a confidence interval, or a conclusion without any risk-of-bias caveat, overstates certainty relative to what the Methods and Limitations sections actually support.
• Conclusions that outrun the evidence. If the Limitations section flags substantial risk of bias or heterogeneity, the Conclusions section needs to be hedged accordingly, not stated as a firm recommendation.
• Treating the abstract as a teaser. Unlike a narrative introduction, a structured systematic review abstract must be able to stand alone — a reader or indexing database should be able to judge the review’s scope, method, and main finding without opening the full text.

Frequently asked questions

How long should a systematic review abstract be?

There is no single required length. ICMJE does not mandate a word count and leaves it to each journal; most journals that require structured abstracts for systematic reviews set a limit somewhere between roughly 250 and 350 words. Always check the target journal’s instructions for authors before finalizing length.

What’s the difference between a systematic review abstract and a regular structured abstract?

Both use labeled sections, but a systematic review abstract additionally has to report search sources, eligibility criteria, risk-of-bias methods, and (per PRISMA 2020 for Abstracts) a registration number and funding statement — elements a single-study abstract doesn’t need because it isn’t synthesizing a body of literature. See the general research paper abstract worked example for comparison.

Do systematic review abstracts need a PROSPERO number?

PRISMA 2020 for Abstracts asks authors to provide the register name and registration number if the review was registered. Registration itself (commonly with PROSPERO) isn’t universally mandatory, but many journals and funders now expect it, and if a review is registered, PRISMA expects that number to appear in the abstract.

Is a systematic review abstract the same as a graphical abstract?

No. A graphical abstract is a visual, single-figure summary some journals request alongside the text abstract; it doesn’t replace the structured text abstract PRISMA and ICMJE describe, and it isn’t governed by the same checklist.

Where can I read more about PRISMA methodology generally?

See CASRAI’s PRISMA and Systematic Review Methodology guide for the full 27-item checklist and flow diagram requirements that govern the rest of the paper, and the systematic review dictionary term for the underlying definition. For the equivalent worked-example treatment of other manuscript types, see the research paper abstract worked example and case report abstract (CARE guidelines) worked example.