Scoping review

What a scoping review is

A scoping review is a type of evidence synthesis that aims to map the literature on a broad topic. Where a systematic review asks a focused question about the effect of an intervention or the accuracy of a test and weighs the quality of studies to give an answer, a scoping review asks what evidence exists, what kinds of studies have been done, how concepts have been defined, what the research gaps are and how the field might be organized. It is exploratory by design. The output is a description of the evidence, in tables, charts and a narrative, and not a pooled estimate or a judgment of effectiveness.

The approach became widely used in health and social care, education and many other fields after a framework by Arksey and O'Malley in 2005, and has since been refined by others, with guidance from the Joanna Briggs Institute and a reporting guideline, PRISMA-ScR. Although it is called a review, the methods are systematic in the sense that the search and selection are explicit and reproducible. See the guide on choosing a review type.

When to choose a scoping review

A scoping review suits several situations. It is useful when a topic is broad or complex and has not been comprehensively reviewed, so that a focused systematic review is premature. It helps to examine how research on a subject has been conducted, including the designs, populations, concepts and outcomes used. It can clarify working definitions and the boundaries of a concept, which is often needed in fields with inconsistent terminology. It can identify gaps that justify further primary research or a later systematic review. And it can be used to establish whether a systematic review is feasible, by showing whether enough studies of a suitable kind exist. It can also precede a systematic review as its first stage.

It is not the right choice when the aim is to know whether an intervention works, how accurate a test is, or what the pooled size of an association is, since it does not assess the quality of evidence or estimate effects. A scoping review should not be chosen only because it seems quicker: its breadth often makes it lengthy, and it has methodological requirements of its own. Guidance by Munn and colleagues sets out questions to help decide between the two designs. See systematic review.

The main frameworks

Arksey and O'Malley

The original framework proposed five stages: identify the research question, identify relevant studies, select studies, chart the data, and collate, summarize and report results, with an optional sixth stage of consultation with stakeholders.

Levac and colleagues

Recommendations to strengthen the framework: clarify and link the purpose to the question, balance feasibility with breadth, use an iterative team approach to selection and charting, incorporate a numerical summary and a thematic analysis, and make consultation a required stage.

Joanna Briggs Institute

A methodology with detailed guidance on the title, the question using the population, concept and context framework, inclusion criteria, protocol, search strategy in three steps, selection, extraction, analysis and presentation, with updated guidance in 2020.

PRISMA-ScR

A reporting guideline of 20 essential items and two optional items, an extension of PRISMA specific to scoping reviews.

Reviews should say which framework they followed and where they departed from it.

Conducting a scoping review

1. Define the question and objectives

   The question is broad but bounded, and is often framed with the population, concept and context. The objectives say what the map is for. A protocol is written and may be registered or published.

2. Search

   The search is comprehensive and includes published and, often, grey literature. The strategy is developed iteratively, tested and documented.

3. Select

   Records are screened against criteria that are explicit and may be refined as the team becomes familiar with the literature, with the changes documented. At least two reviewers are recommended.

4. Chart the data

   A charting form records key characteristics of each source, such as authors, year, country, design, population, concept, methods and key findings. The form is piloted and refined.

5. Collate and summarize

   Results are tabulated and presented as a numerical summary of the extent and nature of the studies and a narrative or thematic account, with visual maps.

6. Consult

   Stakeholders, such as practitioners, patients or policy makers, may be asked to comment on the findings and add perspective.

Charting and presenting the map

Charting is the equivalent of data extraction, with the difference that the data are broader and descriptive. The form is designed around the objectives and refined as the work proceeds, because the review is iterative. Typical fields include the year, country, setting, design and methods, population characteristics, the concepts or interventions studied, the outcomes measured, and the main findings relevant to the objectives. There is no assessment of effect size, since the aim is to describe.

The map is presented in several ways. Numerical summaries show the distribution of studies by year, country, design and topic, so that trends and concentrations are visible. Tables describe the characteristics of each source. Evidence maps and heat maps cross two dimensions, such as interventions against outcomes, to show where studies exist and where there are none. A narrative describes themes, how concepts are defined and used, and the gaps. The conclusions concern the state of the evidence and the implications for research, practice or policy, without claims about effectiveness.

Framing the question with population, concept and context

Scoping reviews often use population, concept and context in place of PICO, because the aim is to map and not to compare. The population is the group of interest, the concept is the central phenomenon that the review will map, which may be an intervention, an outcome, a phenomenon or a term whose meaning is unclear, and the context is the setting, culture or circumstance. A question such as what is known about community-based programs for loneliness among older adults is framed by a population of older adults, a concept of programs to reduce loneliness, and a context of community settings. The framing guides the eligibility criteria and the search, and it sets the boundary of the map. A question that is too broad produces an unmanageable review, and one that is too narrow may be better suited to a systematic review, so the framing is often refined after a preliminary search.

Searching in a scoping review

The search is as systematic as for a systematic review, and it is often broader in sources. Guidance describes a three-step approach: an initial limited search in a couple of databases to find relevant terms, a full search across all chosen sources using all identified terms, and a search of the reference lists of included sources. Grey literature, such as reports, theses, policy documents and organizational websites, is commonly included, since the topics of scoping reviews often have much of their material outside journals. Because the criteria may be refined iteratively, the search may be rerun, and each version is recorded with its date. The strategy for at least one database is reported in full, and the search is documented so that it can be repeated.

Analysis and synthesis

Analysis in a scoping review is descriptive. A numerical summary counts the sources by year, country, design, population and the concepts studied, and a narrative or thematic analysis groups the content of the sources into categories, for example by type of intervention, by definition used, by outcome measured or by gap. Qualitative content analysis can be used to categorize findings. The grouping should be driven by the objectives and not by what is easiest to count. Because there is no weighing of evidence, claims are limited to what the map shows: that evidence exists or not on a question, how it is distributed, and how concepts are used. The discussion considers what the map implies for practice, policy and research, including where a systematic review would now be feasible.

Quality appraisal

Scoping reviews do not usually assess the risk of bias or methodological quality of included sources, because their purpose is to describe the extent of evidence and not to judge it. This is a defining difference from systematic reviews and a frequent source of criticism, since readers may take the existence of studies as evidence of their reliability. The review should therefore be explicit that inclusion does not imply quality. Some reviews do appraise sources when it helps their aims, and PRISMA-ScR allows reporting of this. Where appraisal is done, it should be justified and its use stated. In every case, the conclusions should not be phrased in a way that treats the volume of research as evidence of effect.

Consultation

Consultation with people who use or are affected by the research, such as clinicians, patients, policy makers and community members, is a recommended stage. It can help to refine the question and the criteria, to identify sources the search missed, to interpret the findings, and to shape the implications. It adds relevance, and it requires planning: who is consulted, how, and how their input is used and reported. Consultation is not a substitute for the review's own methods, and it should be described openly, including how views that differ from the evidence were handled.

Reporting

PRISMA-ScR specifies what a scoping review should report. The items include the rationale and objectives, the existence and location of a protocol, the eligibility criteria and the rationale for them, the information sources and the full search strategy of at least one database, the selection process, the data charting process and items, the synthesis of results, the characteristics of the sources and the results of individual sources of evidence, and the summary of evidence in relation to the objectives, limitations and conclusions. A flow diagram shows the selection. Protocols can be registered on open platforms, since PROSPERO does not accept scoping reviews. See PRISMA extensions.

Limitations and common problems

Scoping reviews are criticized when they are used as an easier route to a systematic review, when the question is so broad that the review cannot be completed in a meaningful way, or when the findings are interpreted as showing effectiveness. The breadth of the question often produces large numbers of records, and the iterative changes in criteria and charting need careful documentation to remain transparent. Quality is not assessed, so the conclusions are limited. Reviews vary in methods, and conventions are still developing. A scoping review is a legitimate and useful design when its purpose is mapping, and a misleading one when used to answer questions that need effect estimates.

How we can help

Frequently asked questions