Systematic Review

A systematic review is a literature review that uses an explicit, pre-specified, and reproducible protocol to identify, select, appraise, and summarise all the studies relevant to a focused research question. Its defining feature is method: by deciding in advance how studies will be sought and judged, it aims to minimise the bias and arbitrariness of a traditional narrative review.

Definition

A systematic review is a review of a clearly formulated question that uses systematic and explicit methods to identify, select, and critically appraise relevant research, and to collect and analyse data from the studies that are included in the review.

Scope

This entry covers what distinguishes a systematic review from an ordinary literature review: a registered or pre-specified protocol, a focused question, comprehensive searching, duplicate study selection against explicit eligibility criteria, risk-of-bias appraisal of included studies, and transparent reporting. It treats the systematic review as a method of evidence synthesis, not as clinical advice.

Core questions

• What does the totality of relevant evidence say about a single, clearly defined question?
• Which studies meet the pre-specified eligibility criteria, and how trustworthy is each one?
• How can the search, selection, and appraisal be made transparent and reproducible?

Key concepts

• Pre-specified protocol and registration
• Focused question (e.g. PICO framing)
• Comprehensive, reproducible search
• Explicit eligibility criteria
• Duplicate study selection
• Risk-of-bias appraisal
• Transparent reporting (PRISMA)

Mechanisms

The systematic review controls bias by fixing its methods before the results are known. A focused question (often structured as population, intervention, comparator, outcome) defines what counts as relevant. A comprehensive search across multiple databases and sources aims to retrieve all eligible studies, reducing the chance that the conclusion is driven by a convenient subset. Two reviewers independently screen and select studies against explicit criteria, limiting selective inclusion, and appraise each study's risk of bias so that the synthesis weighs trustworthiness. Reporting standards (PRISMA) make every step inspectable, and protocol registration (PRISMA-P) guards against outcome switching and post-hoc changes. A meta-analysis may then pool the results, but the systematic review is the method whether or not the data are pooled.

Clinical relevance

Systematic reviews underpin clinical practice guidelines and health-technology assessment and are widely treated as a high tier of evidence when well conducted. Knowing how to read one critically, including its search, selection, and appraisal methods, is a core evidence-appraisal skill. This entry explains how such reviews are produced and judged; it is not guidance for managing an individual patient.

Epidemiology

Systematic reviews are produced across medicine, public health, nursing, and the social sciences, and organisations such as the Cochrane Collaboration coordinate large numbers of them in health care. PRISMA reporting is now expected by most biomedical journals, and protocol registries such as PROSPERO record planned reviews. Output has grown rapidly, prompting attention to quality and redundancy.

Evidence & guidelines

Reporting is governed by the PRISMA 2020 statement (Page et al., 2021) and, for protocols, PRISMA-P (Moher et al., 2015); the 2009 PRISMA statement and its explanation-and-elaboration paper remain widely cited foundations. These are reporting and conduct standards, not treatment recommendations.

History

Modern systematic reviewing grew out of the evidence-based medicine movement and Archie Cochrane's 1970s critique that medicine lacked systematic summaries of its trials. The Cochrane Collaboration, founded in 1993, institutionalised the practice. Reporting standards evolved from the QUOROM statement to PRISMA in 2009 (Moher et al.) and its 2020 update (Page et al.), while PRISMA-P (2015) addressed the prior registration of review protocols.

Debates

How comprehensive must the search be?

Reviewers weigh the ideal of retrieving every eligible study against the practical limits of time, databases, grey literature, and languages; incomplete searching can bias conclusions, yet exhaustive searching is costly and is itself a methodological judgement.

Key figures

• Archie Cochrane
• Iain Chalmers
• David Moher
• Alessandro Liberati
• David Sackett

Seminal works

• moher-2009-prisma
• page-2021-prisma
• sackett-1996-ebm

Frequently asked questions

How does a systematic review differ from a narrative literature review?

A systematic review pre-specifies and reports its methods, including how studies were searched for, selected, and appraised, so the process can be reproduced and scrutinised. A narrative review usually has no such protocol and may reflect the author's selective choice of studies.

Does a systematic review always include a meta-analysis?

No. A systematic review is defined by its method, not by statistical pooling. When included studies are too few or too dissimilar to combine, the review synthesises them qualitatively instead of computing a pooled estimate.