A project in outline
A meta-analysis or systematic review is a sequence of decisions that depend on one another. The question determines the eligibility criteria, the criteria determine the search, the studies found determine which analyses are possible, and the analyses determine what can be reported. For that reason the work is organized in stages, and each stage ends with something you can review before the next begins.
Projects do not always begin at the beginning. Many researchers arrive with a question and no protocol. Others have already screened studies and extracted data and need the analysis. Some have a complete manuscript and need responses to reviewers. The scoping stage establishes where your project stands and which of the stages below apply, so that you do not pay for work that is already done and do not skip work that would later be questioned.
The sections that follow describe the stages in the order a full review project would meet them. The services behind them are described under systematic review, meta-analysis and publication support.
What helps before you get in touch
You do not need a finished plan to start a conversation, but a few facts shorten the scoping stage and make the first reply more useful.
- The question, in your own words. A sentence or two is enough. If you can name the population, the intervention or exposure, the comparison and the outcome, so much the better.
- The kind of project. A systematic review, a meta-analysis, a network meta-analysis, a scoping review, or a stage within one. It is fine to be unsure.
- What already exists. A protocol, search results, a list of included studies, an extraction spreadsheet, a draft, or reviewer comments.
- The target journal and deadline. Journals differ in the reporting guideline they require, the word limit they set and their view of different review types, so the target shapes the plan.
- Existing reviews you know of. If a recent review covers the same question, the new project needs a clear reason to exist, and it is better to know that at the start.
The contact form asks for these items. It does not accept file uploads, so describe what you have and we will explain how to share it.
Stage 1: Enquiry and scoping
The first stage is a review of your enquiry and, where needed, a short exchange of questions. The aim is to understand the question, the evidence base, the state of the project and the constraints, and to judge whether the proposed design suits them.
Two judgments matter most here. The first is feasibility: whether there are likely to be enough comparable studies, with usable data, for the planned synthesis. Pooling cannot be promised before the studies are known, and a project that looks like a meta-analysis may turn out to need a narrative or scoping approach. The second is novelty: whether existing reviews already answer the question. Where a recent, high-quality review exists, the plan may need to change to an update, a new outcome or a different population.
The outcome of this stage is a written scope. It states which stages will be done, what will be delivered at each, what assumptions it rests on, and what would change the plan. You can use it to decide whether to proceed.
Stage 2: Agreement
Once the scope is clear, the terms are agreed in writing before work starts. The agreement covers the scope and deliverables, the schedule and its dependencies, the fee and when it is due, the number of revision rounds included, and how authorship and acknowledgment will be handled. The general terms are set out in the terms of service.
Authorship deserves attention at this point and not at submission. Support with analysis, writing and editing is legitimate and common, and it is described openly in the research integrity and authorship statement. The researchers who conceive and direct a study remain responsible for its content and conclusions, and contributions that do not meet the criteria for authorship are acknowledged. Agreeing this early avoids difficulty later.
Stage 3: Protocol and analysis plan
A protocol records the review question, the eligibility criteria, the sources to be searched, how studies will be screened, what data will be extracted, how risk of bias will be assessed, and how results will be synthesized. Writing it before the search is run limits the freedom to alter criteria or analyses after the results are seen, which is the main safeguard against selective reporting. The PRISMA-P guideline lists the items a protocol should contain.
Where a suitable registry exists, the protocol is registered. PROSPERO accepts systematic reviews with a health-related outcome, and other platforms can be used for reviews outside its scope. For a meta-analysis, the protocol also includes the analysis plan: the effect measure, the model, the planned subgroup and sensitivity analyses, and how dependent data and missing data will be handled. The reasoning is covered in more detail under protocol development and registration.
You review and approve the protocol. Changes made after this point are recorded as deviations with a reason, and they are reported in the final manuscript.
Stage 4: Searching and selecting studies
The search strategy translates the question into the vocabulary of each database, combining controlled subject headings with free-text terms and checking the result against a set of studies already known to be relevant. The databases searched depend on the field. A health review would normally include MEDLINE, Embase and the Cochrane trials register, while a psychology, education or management review would draw on different sources. Trial registries and, where appropriate, grey literature are searched to reduce the risk that unpublished results are missed.
The searches are saved with their dates and the number of records retrieved, so that they can be reported and, if necessary, repeated. Records are de-duplicated and screened against the eligibility criteria, typically in two steps (titles and abstracts, then full texts), and the reasons for excluding full texts are logged. The result is the flow diagram that PRISMA 2020 requires. See search strategy development and the guide to building a search strategy.
Stage 5: Data extraction and appraisal
Data are extracted onto a form that was piloted on a few studies before use. The form captures study characteristics, participant details, interventions or exposures, outcome definitions and the numerical results needed for the analysis. Where the protocol calls for it, two people extract independently and resolve disagreements, because transcription and unit errors are among the commonest causes of wrong pooled estimates.
Each study is also appraised for risk of bias with a tool suited to its design, such as RoB 2 for randomized trials, ROBINS-I for non-randomized studies of interventions, or QUADAS-2 for diagnostic accuracy studies. The judgments feed into the interpretation of the results and, for health outcomes, into a GRADE assessment of the certainty of the evidence. See data extraction and coding and risk-of-bias and certainty assessment.
Stage 6: Statistical analysis
The analyses are run as specified in the plan. They typically include the pooled estimate with its confidence interval, heterogeneity statistics, a prediction interval, the planned subgroup analyses or meta-regression, sensitivity analyses, and an assessment of small-study effects when enough studies are available. Each result is reviewed for plausibility, and key results are re-computed by a second method to detect coding errors.
You receive the results with the figures and a written description of the methods, together with the analysis code so that the work can be reproduced. If the data turn out to be unsuitable for the planned approach, for example because outcomes cannot be reconciled, that is reported to you and the alternatives are discussed before anything is changed. The statistical approach is described in detail on the meta-analysis service page.
Stage 7: Reporting and manuscript
Results are reported against the guideline that applies to the review, usually PRISMA 2020 and, for meta-analyses of observational studies, MOOSE. A completed checklist, the flow diagram, the tables and the figures are prepared for the manuscript. Where the agreed scope includes manuscript preparation, the full paper is drafted or revised on the basis of the analysis and your interpretation of the findings, with a reference list checked against the sources.
You review the draft and decide what the paper says. Interpretation, conclusions and the discussion of implications are yours to judge, and changes you request are made within the agreed revision allowance. The manuscript is prepared for your submission, not published or submitted in your name without your approval.
Stage 8: Submission and revision
Where submission support is part of the scope, it begins with the choice of journal. The considerations are the journal's scope, its stated openness to evidence-synthesis papers, the reporting guideline it requires, its indexing and its peer-review model. Checks on whether a journal is reputable are part of this, because predatory journals are a real risk for authors. The manuscript is then formatted to the instructions for authors and the cover letter and supplementary files are prepared.
No journal can be promised to accept a paper. After submission, the reviewers' and editor's comments are reviewed with you, and a point-by-point response is drafted. Requests for further analysis are carried out and reported, and the revised manuscript is checked against the response before it is returned. See peer-review revision support.
Communication, files and version control
Each project has one agreed channel for communication, so that decisions are recorded in one place. Files are exchanged in a way that preserves version history, and each deliverable is labeled with the date and the version of the protocol or analysis plan it follows. Data and manuscripts shared with us are used only for your project and are not reused. How personal data in enquiries is handled is described in the privacy policy.
Where results depend on a decision, such as which outcome to prioritize when several are reported, the decision is put to you in writing with the consequences described, and the answer is recorded in the analysis log.
What affects the schedule
No single timeline fits every project, and a fixed promise made before the studies are known would be unreliable. The factors that matter most are listed here.
- The number of studies and the effort to extract from them. Studies that report results fully are quick to extract. Studies that require derivation or contact with authors are slower.
- The number of outcomes and subgroup questions. Each outcome multiplies the extraction, appraisal and analysis work.
- Whether searching and screening are included. A review that starts from the search takes longer than one that starts from a set of included studies.
- The speed of your reviews and decisions. The schedule depends on prompt feedback at each stage.
- Journal requirements. Some journals require registration, specific checklists or supplementary files that add steps.
- The peer-review process. Review times after submission are set by the journal and are outside anyone's control.
After scoping, a schedule is proposed with its dependencies stated. If you have a fixed deadline, such as a thesis submission or a grant report, say so at the start so that the scope can be set accordingly.
Revisions, changes in scope and exclusions
The agreement states how many rounds of revision are included. Changes that fall within the agreed scope, such as correcting an error or clarifying a passage, are handled within it. Changes that alter the scope, such as adding an outcome, widening the eligibility criteria after screening has begun, or repeating a search, are discussed first, because they affect the schedule and the cost and, if made after results are seen, they affect how the findings must be reported.
Some things fall outside every project. We do not provide clinical advice or advice about the care of individual patients. We do not alter data, select results to reach a particular conclusion, or write references that cannot be verified. We do not guarantee acceptance by a journal, publication in a specific journal, or a particular statistical result. These boundaries are set out further in the research integrity statement and the disclaimer.
Confidentiality and data handling
Unpublished manuscripts, data and ideas are treated as confidential. Access is limited to the people working on your project, and materials are not used for any other purpose. If a project involves data from human participants, the data you share should already be de-identified, and you remain responsible for confirming that you have the right to share them. Tools that send material to third parties are not used with confidential content without your agreement. The position on automated tools is set out in the AI-use policy.
Three typical engagements
- From question to manuscript
- The project begins with scoping and a protocol, runs through search, screening, extraction, appraisal and analysis, and ends with a manuscript and submission support. This is the longest route and involves all of the stages above.
- From extracted data to results
- You have screened studies and extracted data. The work starts with a data audit and an analysis plan, followed by the analysis, figures, and the methods and results text. See the meta-analysis service.
- From reviewer comments to revision
- You have submitted a manuscript and received comments on the statistics or the reporting. The work starts by reading the comments and the manuscript, then re-analysis where it is requested, and a point-by-point response.
Frequently asked questions
Do I need a finished protocol before contacting you?
No. You can start with a research question, and the protocol is one of the early stages of a full project. A short description of the question, the kind of project and the target journal is enough for the first reply.
Can you start from the analysis if I have already extracted the data?
Yes. A project can begin at any stage. Extracted data are checked before analysis, because errors at that point carry through to the pooled estimates.
Will you guarantee that my paper is accepted?
No. No honest provider can. Acceptance depends on the journal, the editors and the reviewers. What can be done is to make the methods sound, the reporting complete and the submission well prepared.
Who decides what the conclusions say?
The authors. Support with analysis, writing and editing does not change who is responsible for the content of a paper and its interpretation. Contributions that do not meet the criteria for authorship are acknowledged.
How long will a project take?
It depends on the number of studies, the number of outcomes, which stages are included and how quickly decisions are made. After scoping, a schedule is proposed together with the dependencies it rests on.
Can I share my data and manuscript securely?
The enquiry form does not accept files. After you describe your project, the way of sharing materials is agreed, and they are treated as confidential.
References
- Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71. doi:10.1136/bmj.n71
- Moher D, Shamseer L, Clarke M, et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev. 2015;4:1.
- Higgins JPT, Thomas J, Chandler J, et al., editors. Cochrane Handbook for Systematic Reviews of Interventions. Cochrane; current version available at training.cochrane.org/handbook.
- Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. JAMA. 2000;283(15):2008-2012.
- Grudniewicz A, Moher D, Cobey KD, et al. Predatory journals: no definition, no defence. Nature. 2019;576:210-212.
- International Committee of Medical Journal Editors. Recommendations for the conduct, reporting, editing, and publication of scholarly work in medical journals. Available at icmje.org.