The main types of literature reviews are the narrative review, the systematic review, the scoping review, the rapid review, the umbrella review, and the integrative review, and they differ chiefly by how reproducible their method is and what question they answer. Choosing the right one is less about prestige and more about matching the review’s rigour to your aim, your timeline, and the maturity of the evidence.
How to tell the types apart
Rather than memorising labels, sort any review by three questions: is the method reproducible, does it appraise study quality, and does it answer a focused question or map a field? Those three axes separate a casual summary from a guideline-grade synthesis. A review can be fast and broad, or slow and definitive, but it rarely manages both. The sections below place each common type on those axes so you can pick deliberately.
The reproducible, question-answering reviews
Systematic review
The most rigorous format, a systematic review answers one pre-specified question through a documented search, duplicate screening against explicit eligibility criteria, formal risk-of-bias assessment, and a transparent synthesis. It is the benchmark when you need a defensible, repeatable answer.
Meta-analysis
Often described as a type of review, a meta-analysis is really the statistical step inside a systematic review that pools comparable results into a single weighted estimate. It is only appropriate when studies are similar enough to combine, judged partly by heterogeneity.
Umbrella review
An umbrella review sits one level higher again: it systematically reviews existing systematic reviews on a topic, useful when a field already has several reviews that need to be reconciled into a single overview.
The mapping and exploratory reviews
Scoping review
A scoping review follows a reproducible process but aims to chart the breadth of a field, what evidence exists and where the gaps are, rather than to appraise quality or pool results. It suits emerging or heterogeneous topics, as we cover in detail in the systematic versus scoping comparison.
Rapid review
A rapid review streamlines the systematic method to deliver an answer faster, for instance by limiting databases or having a single reviewer screen with a second checking a sample. It trades some rigour for speed and is best when a decision cannot wait for a full review.
Integrative review
An integrative review deliberately combines diverse study designs, including qualitative and theoretical work, to build a broader understanding of a concept. It is more structured than a narrative review but less prescriptive than a systematic one.
The traditional narrative review
The narrative review, sometimes called a traditional or literature review, summarises a field selectively at the author’s discretion. It is fast and flexible, ideal for teaching, background sections, and conceptual essays, but it is not reproducible and should not be presented as definitive evidence. The full contrast is laid out in our guide to the difference between systematic and narrative reviews.
The three axes that separate the types
Instead of memorising six labels, place any review on three axes and the right name falls out. The axes are reproducibility of method, whether study quality is formally appraised, and whether the aim is to answer a focused question or to map a field. The common types line up like this:
- Systematic review: reproducible, appraises quality, answers a focused question. The full-rigour benchmark.
- Scoping review: reproducible, usually does not appraise quality, maps breadth rather than answering an effect question.
- Rapid review: partly reproducible with documented shortcuts, light or selective appraisal, answers a focused question under time pressure.
- Umbrella review: reproducible, appraises the included reviews, answers a question by synthesising existing reviews rather than primary studies.
- Integrative review: semi-structured, variable appraisal, combines diverse designs to build a concept.
- Narrative review: not reproducible, no formal appraisal, summarises a field at the author’s discretion.
The two extremes worth keeping straight are the systematic review and the meta-analysis nested inside it at the rigorous end, and the narrative synthesis used when pooling is unwise as the honest fallback within an otherwise systematic review. A review can move along an axis, but it cannot be both maximally fast and maximally definitive at once.
Common mistakes when choosing a review type
The expensive errors are nearly always made at the planning stage, before a single search line is written:
- Calling a narrative review systematic. Labelling a selective summary as “systematic” without a documented search, duplicate screening, or appraisal misleads readers and is spotted immediately at peer review.
- Choosing a systematic review to map a broad, immature field. When the literature is sparse or wildly heterogeneous, a scoping review answers the real need, “what exists?”, far better than a forced effect question.
- Reaching for a rapid review when the decision can wait. The speed shortcuts trade away rigour, so they are only justified when a decision genuinely cannot wait for a full review.
- Switching type mid-project. Deciding after the search that the question was really a scoping one wastes the work, because the search, criteria, and appraisal plan differ by type. The choice belongs in the protocol.
Each of these is cheap to avoid up front and costly to fix later, which is why our protocol service settles the review type before any searching begins.
Choosing the right type
Work backward from your need. If a decision-maker or examiner must be able to audit your method, choose a reproducible type and write a protocol first. If your field is broad and you need to map it before committing, a scoping review fits. If you are pressed for time, a rapid review may be enough. And if the work is purely background or conceptual, a narrative review is the leaner option. Whatever you choose, defining the method up front, including the review question, is what keeps the result credible.