What do you think are the most useful methods of ensuring suitable and useful search results when Utilising databases for research?
What is a systematic review?A systematic review is a method of answering a question by finding, appraising and synthesising evidence in a systematic and unbiased way. Show
This guide will take you through the steps involved in conducting a systematic review. Click one of the stages below to see information, tips, resources and support available for each part of the process. Preparation(average timeline - 5 months*) Before you startdetermining if a systematic review is the right fit and an overview of the logistics in conducting a review. Define research questionscoping your topic, types of research questions and writing a focused question for your review to answer. Conducting reviews(average timeline - 10 months*) Perform searchescollecting the evidence, recording your search, and exporting to a reference manager. Screen resultsdetermining which studies to include in the review and methods for reporting on decisions. Write and publish(average timeline - 3 months*) *this is a rough estimate of the time needed to complete each stage, but will vary depending on the complexity of your review. Systematic Review Service CharterThe
Library provides tiered support for researchers conducting systematic reviews. Postgraduate students who conduct systematic reviews as part of their degree can request up to 3 hours of support per systematic review. Postdoctoral researchers and beyond can opt for more extensive support, depending on their requirements and our capacity. For details, please refer to the
service charter. If you are interested in Tier 3 (Premium) support, please submit a request and the relevant team will review it. FeedbackWe are always improving our resources. If you'd like to share your feedback with us, please do so via our feedback form. Before you startIf you haven't conducted a systematic review before, you might be wondering where to start. This section takes you through a series of questions to consider before starting a systematic review.Is my question appropriate for a Systematic Review?The systematic review process works best for a clearly defined question with a narrow scope and a single answer. Traditionally, systematic reviews in health disciplines investigate the effectiveness of an intervention (or compare the effectiveness of two interventions), but the methodology is continually adapting to address different question types and disciplines. A systematic review may not be the best method for your research if your question is too broad or the existing evidence is too sparse or varied. There are other review types that share some of the systematic review's characteristics, namely scoping reviews and systematic literature reviews. Explore differences and similarities of various review types in this comparative table: Types of Reviews. Complete this online module to find out which review type is the most suitable for your research question: Scoping reviews In many cases, the most appropriate review type for a literature review for a PhD thesis would be a scoping review. To learn about scoping review methodology, refer to Additional Reading section on this page. What type of systematic review is the most appropriate for my question?If you are certain that your question can be answered by conducting a systematic review, you need to consider what type of systematic review suits it best. All systematic reviews share the same methodology. However, different types of questions require different considerations. To learn what kind of systematic review is the most applicable to your question, complete this online module Is there an existing systematic review on my topic?Before starting work on a systematic review check if a systematic review on a similar topic has already been published (or commenced). Use the following tools to perform a quick search for your topic (a few main keywords will do), and limit to "reviews" or "systematic reviews" (depending on the interface):
Do I have enough time to complete a systematic review?Systematic reviews take time. The estimated timeline for a Cochrane systematic review is 18 months full time. Doing a systematic review properly can take months, even if you're not conducting a Cochrane-style systematic review. If you expect that time might be an issue, consider conducting a rapid review instead. For more on rapid reviews, refer to Additional reading section on this page. Do I have access to necessary resources?A team is an important factor for a successful systematic review. Your team may include:
If you are planning to perform a meta-analysis, you will also need access to:
Finding collaborators can be tricky. You might consider:
Depending on your topic, you might want to explore professional networks to identify potential collaborators among a range of stakeholders including clinicians, patients, customers, or anyone else whose contribution to the review might be useful.
Systematic Review GuidesMost systematic reviews are currently conducted in fields like health and medicine. Increasingly, systematic reviews are being published in other disciplines such as humanities and education. Throughout this guide, you might need to consult with Cochrane, the Joanna Briggs Institute and the Campbell Collaboration guides for more specific advice. These organisations are responsible for the development of many best practice frameworks and guidelines. The Joanna Briggs Institute reviewer’s manual (2020) Systematic reviews in practice by Crawford, Boyd, & Jonas (2015) Cochrane handbook for systematic reviews of interventions (2019)
How to do a systematic literature review in nursing: a step-by-step guide by Bettany-Saltikov (2012) Doing a systematic review: a student's guide by Boland, Cherry, & Dickson (2014) Campbell Collaboration
Need help?To learn what levels of library support are available for researchers conducting systematic reviews explore the SR Service Charter. The library runs regular workshops to get you started with your literature review. To see upcoming workshops and to register, check the Library Calendar.
Additional Reading
Define research questionA focused research question is imperative for the success of your systematic review. This section provides tips on how to make your research question suitable for a systematic review. Research topic vs a review questionA 'research topic' is the area of study that you are researching, while a 'review question' is the more focused question that your systematic review aims to answer. Depending on your starting point, you may arrive at a review question by taking different routes. If you are working in an Applied Sciences field, or as a clinician, your review question might be informed by a problem or a scenario encountered in the lab or in practice. If you are working towards your PhD, your thesis topic will be much broader than a question for a systematic review. Often, to account for the many topics covered in your thesis, you may need to consider writing a scoping review (for more, see Before you start. For example, your thesis might be about studying a range of non-therapeutic treatments for dementia. In your scoping review you will identify all existing treatments, and in your systematic review you might focus on the effectiveness of only one treatment: Example Research topic and scoping review (broad)
Systematic review question (narrow)
Focussing Research QuestionTo help you focus your research question we suggest exploring different mnemonics. Mnemonics highlight different concepts that you need to focus on. To explore different mnemonics developed for different disciplines refer to the Additional literature section on this page. In health and medicine, mnemonics might differ depending on the type of question you are working on, e.g., aetiology, prognostic, prevalence, etc. You can explore different mnemonics in this article For more information on mnemonics refer to Ready for Health Research Module 2. Why PICO?The most widely used mnemonic for systematic reviews is PICO – Population, Intervention, Comparator, Outcomes. PICO can be used to formulate a focused research question for systematic reviews in any field where the objective is to assess the effectiveness of an intervention. PICO originated from evidence-based practice where it is used to help clinicians identify the most important elements of a question when analysing a patient’s condition and suggesting the best treatment. For example, if your research topic was to explore a range of non-therapeutic treatments for depression, you could use PICO to concentrate on the specific conditions, treatments, and their outcomes: Example
You can learn about evidence-based practice and the use of PICO by exploring this EBP subject guide. For more information on how to use PICO refer to Ready for Health Research Module 2. Limitations of PICOWhen applying PICO to your research topic, remember:
For more information on how to translate main concepts into searchable concepts, refer to Ready for Health Research Module 2. Develop and register protocolWriting a protocol is an important step in minimising potential bias in the systematic review process. This section provides tips on how to prepare and register a protocol, and define eligibility criteria. Why do I need to write a protocol?A protocol specifies the question a systematic review will attempt to answer and the methods that will be used to answer it. Having a protocol is essential because:
Preparing a protocolThere are many reporting standards available which outline the essential information you need to provide in a protocol. Requirements will change depending on where you publish the protocol. Some of these templates include: Checklist - PRISMA-P Templates: Campbell Collaboration Template text for reporting search methods - Cochrane SYRCLE Amending a protocolSometimes reviewers need to change their original protocol after a review has commenced. Any changes made should be reported and explained in an amendment to the protocol, as well as in the systematic review itself (Gough, Oliver, & Thomas, 2017). All registers have procedures for submitting amendments. Your search strategy needs to be included as part of your protocol Refer to Plan search strategy and select databases section for more details. Why do I need to register a protocol?You must have a protocol even if you are not planning to register it. Registering a protocol is highly recommended for the following reasons:
Registering a protocolYou have a range of platforms to choose from when deciding where to register a protocol. Many of the platforms will have specific conditions and limitations that you need to be aware of; refer to the Note section for useful tips. PROSPERO
Note: PROSPERO does not register scoping reviews and does not accept submissions from students. Cochrane Reviews Note: when you register your protocol with Cochrane you commit to the Cochrane-style systematic review. Your topic would need to be approved by one of the 52 Cochrane Review Groups. You will be asked to update your review every two years. Joanna Briggs Institute Note: You can register your systematic review title with JBI, including scoping review titles, if at least one of the authors belongs to an "affiliated entity". The JBI journal JBI Evidence Synthesis charges US$1000 upon acceptance of full protocols if none of the authors of the submission are members of the JBI Collaboration. Campbell Collaboration Note: Campbell will not publish protocols of reviews that are to be published elsewhere. Alternative PlatformsYou can submit the review protocol as a preprint on one of the following platforms that will provide you with a time stamp indicating that the protocol was created prior to starting the review: Sydney eScholarship (SES) Open Science Framework Medrxiv Research Square FigShare Collaborative Approach to Meta-Analysis and Review of Animal Data from Experimental Studies Eligibility (inclusion and exclusion) criteriaEligibility criteria, also referred to as inclusion and exclusion criteria, are essential for systematic reviews. They are defined at the start of the systematic review process (a priori) to help you select the most reliable evidence as part of the screening process. Eligibility criteria determine which studies will be considered eligible for inclusion in the final analysis. Because your eligibility criteria must not change in the process of conduction the review, define these criteria carefully and in a way that allows enough flexibility without introducing additional bias. Eligibility criteria constitute part of the protocol. Some criteria may include:
Eligibility criteria may include features that you want either in or out of the studies considered relevant. You can choose the most convenient way of formulating a criterion either as an inclusion, or as an exclusion. You do not need to match each criterion with its opposite, i.e. include papers in English and exclude papers in all other languages. Inclusion criteria examples
Exclusion criteria examples
Each additional criterion introduces a potential bias into your research. It's important to justify each criterion in the methodology section of your systematic review. Different reporting guidelines may require different criteria.
Need help?Preparing a protocol requires the entire systematic review team’s effort. Make sure that you consulted subject experts such as your supervisor. Honours students, Higher Degree Research students and staff members can consult an Academic Liaison Librarian for assistance with search strategies.
Additional Reading
Plan search strategy and select databasesSystematic reviews aim to evaluate all relevant information about a topic to answer a question. This section provides tips on what constitutes a comprehensive search and which information sources you need to consider. What constitutes a comprehensive search?A comprehensive search has three distinct characteristics:
How do I plan a comprehensive search?To construct a comprehensive search, you must:
We suggest you create a concept table to capture all the required information for a comprehensive search. Download this template to keep track of your work. Where do I search?Each database provides access to different collections of resources, so you’ll need to think about which databases contain the information you’re looking for. If you don't search enough sources, you can miss out on crucial studies. Depending on your review question, you need to consider searching both multidisciplinary and subject-specific databases.
A balance of both database types is required in your search strategy. Cochrane considers Medline (via either Ovid or PubMed), Embase, and CENTRAL to be the most important sources of evidence to search when conducting a systematic review in biomedical disciplines. Health and Medicine DatabasesTo learn about the main databases in Health and Medicine, download this document Where not to searchCan I search OVID? Should I search PubMed or Medline or both? Should I search Google Scholar? Grey literatureGrey literature refers to research material that hasn't been published commercially like a book or a journal article has. The challenge lies in locating it, as it can be either published or
unpublished. Examples of published grey literature include:
Examples of unpublished grey literature include:
Depending on your subject, it might be appropriate for you to include grey literature as evidence
because it provides data which otherwise wouldn't be available through journal articles. Including grey literature also helps reduce publication bias. Search filtersSearch filters are validated methods for finding relevant research in health databases. Filters are commonly used to limit searches to a particular study type, or to find articles on a specific subject. They can be used to enhance your search, limit your search results to only relevant study types, or for identifying additional search terms. It is not recommended to combine multiple study design filters – where your eligibility criteria includes multiple study designs, it is instead recommended not to use a study design filter. Study type filters Cochrane randomised controlled trial filters McMaster University Hedges Project contains filters for the following databases:
Search filters (Scottish Intercollegiate Guidelines Network) PubMed Clinical Queries Search filters (ISSG) CINAHL Clinical Queries Subject area filters CareSearch Australia: find evidence-based literature via PubMed for palliative care and related topics Integrated care: search integrated care literature via PubMed Aboriginal and Torres Strait Islander Health: search for literature about ATSI populations and health issues via PubMed Search Blocks (Biomedische Informatie): add 'blocks' to your database searches to limit them by topic SYRCLE Filter for Animal Studies: designed to limit a search to relevant animal studies
Need help?Academic Liaison Librarians can help Honours students, Higher Degree Research students, and staff plan search strategies. Don’t forget to bring your concept table to your appointment. While we can suggest some resources to help you search for grey literature, actual searching is out of scope in our support for systematic reviews.
Additional Reading
̏ Perform searchesComprehensive searching is a complex skill that is refined through deliberate practice. This section offers tips on how to execute comprehensive searches effectively and efficiently across multiple databases.Where can I get help with searching specific databases?Before you start searching in a database, check the following:
In a database with a thesaurus, you will need to search for subject terms as well as for keywords. Syntax are search operators which can be used to modify and improve the results of keyword searches. They'll vary according to the database. See Flinders University's database syntax guide for a comprehensive table of operators used across different databases. To learn more about specific databases in Medicine and Health, their subject headings, syntax, and search techniques, enrol in Ready for Health Research and complete Modules 3-4. Below you can find quick database search guides for specific databases: Database search guidesHow can I run the same search in all selected databases?Because databases use different syntax and subject headings, you'll need to adapt your strategy to fit each database you search. See Flinders University's guide to translating searches from Medline to other databases like CINAHL, PsycINFO, and Scopus. On Ovid, you are NOT advised to:
What is the right number of search results for a systematic review?There is no magic number of search results that a systematic review search should aim for. The number of retrieved results will vary depending on the topic and the area of study. Established fields will yield more results, while new fields will yield less. A comprehensive search strategy needs to balance two objectives: being inclusive enough to capture as many relevant results as possible while keeping irrelevant results to a minimum. If your search returns too many results, you might need to:
If your search returns too few results, you might need to:
How do I check the relevance of my search results?A good way to check if your search is working is by creating a small sample of relevant articles (5 to 10) and checking if your search retrieved these articles. You can quickly find the relevant articles for the sample set by doing a quick search in Google or PubMed. Search for your topic in the PubMed PubReMiner tool to check why your search might not be working. Whereas a search in PubMed retrieves relevant articles, a similar search in PubReMiner retrieves metadata assigned to the articles during the indexing process, such as author names, journal titles, and subject headings, and calculates the frequency with which the same entries appear in the records. Identify high frequency subject headings used to index the relevant articles and include them in your Medline search. Alternatively, if the articles in the same topic are indexed with a wide range of low frequency subject headings, consider including keywords that consistently appear in titles and abstracts of the relevant articles. When do I stop searching?When conducting a systematic review, your aim is to search enough databases to be able to state with confidence that the literature has been comprehensively searched. Indicators of saturation include:
How do I document my searches?Most systematic reviews include a PRISMA flow diagram in their methods section, along with the search strategies themselves. Record your searches in a log or a document to refer to them when writing up your methodology. You can use the PRISMA chart generator to help you build a PRISMA flow diagram for your own systematic review. To ensure accurate reporting, keep numbers of which records were obtained from each database. Set up folders in your reference manager to help keep track of this. You can refer to the Cochrane Handbook for tips on which elements of your search need to be documented Other methods of locating relevant studies for your reviewTo ensure that you’ve located all relevant studies, you may consider supplementing your comprehensive search with one or more of the following additional types of searching:
Any studies that were located by means other than a reproducible comprehensive database search (including grey literature) should be recorded in a separate box in your PRISMA diagram (Additional records identified through other sources).
Need help?We recommend completing the Ready for Health Research Canvas course for more help performing database searches for a systematic review. You might benefit from completing the course even if you are not a health and medicine researcher. Honours students, Higher Degree Research students, and staff are recommended to complete the module before making an appointment with an Academic Liaison Librarian.
Additional Reading
Screen ResultsComprehensive searches always retrieve more results than will be included in your final analysis. The screening process allows you to filter out irrelevant results by applying your inclusion and exclusion criteria. This section offers tips on efficient screening.Which tool can be used for screening?There are many collaborative tools for screening, both free and paid, to make the process easier. The Library doesn't currently support any screening tools. Popular paid tools include: Covidence - a tool for screening and data extraction. Recommended by Cochrane. Your first review can be conducted free of charge. Subscription fees apply for additional reviews.
Eppi-Reviewer - a comprehensive tool used for the searching, screening, and synthesis stages of a systematic review. Freely available to Cochrane-affiliated researchers. Free alternatives include: A Covidence free trial - trial allows 2 reviewers to conduct 1 review.
Rayyan QCRI - a web-based tool for collaborative screening. Learns selection patterns to assist with labelling of studies. Abstrackr - a web-based tool for collaborative screening and labelling abstracts for relevance. Picoportal - a systematic review platform. More screening tools (both paid and free) can be found using the Systematic Review Toolbox - a comprehensive catalogue of tools to support systematic reviews. How do I use EndNote for screening?The Library supports EndNote which can be set up for managing search results for a systematic review. Please refer to this subject guide to learn a 10-step process to use EndNote for systematic reviews. How do I export searches from different databases?Before you screen your search results, you will need to export them from each database. You should only start exporting when all your search strategies are finalised, and no further edits are anticipated. You can find instructions on how to export from multiple databases on the EndNote subject guide (look under the Specific Database guides tab). Refer to this table to learn EndNote export limits across different platforms and tips on bulk exporting. How do I screen search results?To control bias and minimise the risk of excluding relevant studies, screening is usually done by two or more independent reviewers. They talk through any disagreements in decision making over study inclusion or exclusion. A standard workflow for this stage might be:
If your resources are limited, you might choose to do the first screening yourself and involve the second reviewer for the second screening when selection criteria are applied more rigorously, and for the full-text analysis. Where can I get full text articles?If using EndNote, you can attach most full text articles automatically using the ‘Find Full Text’ function. For the rest, check Library Search and, when you have found the full text, attach it to your article records in EndNote manually (refer to the EndNote manual for instructions). If you are confident that the library does not have access to the required article, you can request it via Resource Sharing. How can I re-run the search?Many publishers will request that you re-run your search if six months or more lapsed between the last search and the date of review admission for publication. To do so, you can run your saved searches and limit them to the last year of search, and then screen results just for the last year to identify any new relevant articles, and update your PRISMA flowchart accordingly. Refer to this document for instructions on how to use EndNote to manage references when re-running systematic review searches.
Need help?Librarians cannot help you with screening; for assistance, contact your supervisor or a subject expert in your faculty.
Additional Reading
Data extraction and appraisalData appraisal is essential for a systematic review. This section provides tips and suggests tools for efficient data extraction and appraisal. How do I extract data?Your next step is extracting the necessary data from your included studies. Not only will this help you assess the validity of the studies you're including in your review but will also help you prepare for the next stage: analysis and interpretation. Before you begin to extract data from the studies, you need to decide what data elements are required for your analysis. The below table is a sample of items to consider. The full table can be viewed in the Cochrane Handbook.
Once you've decided which items are needed, use a data extraction form as a consistent way to record the information. How do I appraise selected studies?When appraising each study eligible for inclusion in your review, you'll be evaluating whether you can draw conclusions about the effect of the intervention. You can assess this by examining the studies for various types of bias, which can lead to the study overstating or understating the true effect of the intervention. Section 8 of the Cochrane handbook names the following forms of bias:
Appraisal toolsThere are a number of appraisal tools which will help you assess the publications / studies you're including in your review.
How do I avoid incomplete reporting?Incomplete reporting can affect how you assess risk of bias. What's excluded has a bearing on whether you can trust the study as a whole, or if you can only trust specific outcomes. Some ways of detecting whether a study has been reported incompletely include:
After appraisal, you might find there are further studies to exclude from the review based on your eligibility criteria. Make sure you record the reasoning behind exclusion as it will need to be reported.
Need help?It's a good idea to talk to your supervisor or a statistician from the Sydney Informatics Hub about how you plan to conduct your analysis, ideally before you begin data extraction. This ensures you're collecting relevant data to support the analysis you plan to undertake.
Additional ReadingCochrane has a GRADE Approach to rating the certainty of evidence in systematic reviews and creating a ‘Summary of findings’ table.
Analyse and interpretTo answer your research question, you will be analysing and synthesising the extracted data. This section provides tips and suggests tools for efficient data analysis. Your review will lend itself to different methods of analysis:
Narrative synthesis techniques
Why should I conduct a meta-analysis as part of my systematic review?
Cochrane Handbook, Section 10.2 'Introduction to meta-analysis' Despite these advantages, a meta-analysis should only be performed if you have a group of studies which are similar - this similarity is referred to as homogeneity. At the very least, studies should be homogeneous in terms of participants, interventions, and outcomes so the meta-analysis can draw a meaningful conclusion. Any kind of variability among studies is referred to as heterogeneity. There are several kinds - Cochrane names:
It is important to measure and address heterogeneity in your review because it affects the extent to which you can rely on the conclusions you draw. Software NVivo
Need help?Your supervisor or a subject expert in your faculty can assist with analysis and interpretation of your results. If you're an HDR student, there is free statistical advice available through the Sydney Informatics Hub. Book an appointment to talk to a statistician by using the assistance request form.
Additional Reading
Write and publishThis section provides tips and considerations for reporting the findings of your systematic review and deciding where to publish.How should I write up my review?In your write-up, you are not only reporting your findings and answering your review question – you are also reporting on the decisions made and the procedures followed that led to your findings. To ensure that your write-up is complete and transparent, use the PRISMA checklist. Where should I publish my review?Where you publish has a huge effect on the visibility and impact of your systematic review. Here are some tips for finding the right journals to publish in according to the scope of your review and your target audience:
How should I share my research?When you submit your article to a journal for publishing, you often sign over copyright to the publishing journal. This affects how you can share it in the future, and how readers can access it. If it is important for your research to be accessible to clinicians and practitioners, check what sharing options a journal offers (often stated in open access or editorial policies on the journal/publisher website), or investigate an open access publication.
Need help?For assistance with writing, contact your supervisor. In addition to providing online modules on writing, The Learning Hub also runs writing workshops. If you have a question about copyright, contact the Copyright Team. Enquiries about Open Access publishing can be directed to an Academic Liaison Librarian. If you have a question about depositing a copy of your publication into the University repository, contact the Repository Team. The Library runs workshops on Strategic Publishing. Check the Library Calendar to see upcoming workshops and to register.
Additional Reading
What is the most effective way to search a database?Top Ten Search Tips. Use AND to combine keywords and phrases when searching the electronic databases for journal articles. ... . Use truncation (an asterisk) and wildcards (usually a question mark or exclamation point). ... . Find out if the database you're using has a "subject search" option. ... . Use your imagination.. What search strategies does the Library recommend using on the databases?Database Search Strategies. Use keywords instead of full sentences or questions. ... . Combine concepts and use multiple search boxes. ... . Use quotes around phrases. ... . Use synonyms combined by “OR” ... . Use the language of the database. ... . Search for all forms of a keyword. ... . Use database search limits to make your results specific, such as:. What other techniques can I use to develop effective search methods?Search strategy techniques. Choosing search terms.. Searching with keywords.. Searching for exact phrases.. Using truncated and wildcard searches.. Searching with subject headings.. Using Boolean logic.. Citation searching.. What other strategies might you use beside key word searches of databases to find research articles for your literature review?Literature Search Strategies: Main. Internet Search Tips.. Generating Search Terms.. Using Library's Databases to Find Articles.. Using Primary & Review Articles.. Evaluating Resources (Four "R"s). Broadening Searches Using One Article.. Citation & Bibliography Tools.. Borrowing from Another Library.. |