Systematic Review Guidance
You run the review. We check the protocol, the search, the screening and the synthesis at each step.
A systematic literature review is a full study in itself, and its method should let another researcher repeat it and get the same papers. For that, the question has to be decided in advance and written into a protocol. The searches cover several databases and are recorded, and two people screen the records independently. Each study you include has to be appraised for risk of bias with a proper tool. In the synthesis you must write its limits openly also. More and more, journals check whether a review called “systematic” really followed this whole method, and a flow diagram alone does not satisfy them.
Scholars mostly come to us at one of three points. Some need a systematic review as a thesis chapter or first publication and want to plan it properly. Some have already started. They have a few thousand records in a spreadsheet and are unsure how to screen them defensibly. And some have finished screening but are stuck at appraisal or data extraction, or they cannot decide if a meta-analysis is possible at all.
We follow the PRISMA 2020 reporting guideline, which is the 27-item checklist and flow diagram published by Page and colleagues in 2021. For methods we go by guidance from sources like the Cochrane Handbook. Two things you should know at the start itself. A systematic review takes months, and a few weeks will not be enough. Also, a meta-analysis makes sense only when the included studies are similar enough to pool. Pool studies that are not alike and reviewers will likely question the number you get. Doing it alone? Our step-by-step systematic review guide has the method.
Quick answers
- Which parts of a systematic review do you check?
Any of seven stages. First is your question and PRISMA-P protocol, which we review. Then the search strings you write for each database, with your search log, and after that training for your screening in Rayyan or Covidence. Next we check your data extraction and risk-of-bias appraisal, and then help with the synthesis or meta-analysis you run. We also do a PRISMA 2020 reporting check, where you fill the checklist with page numbers and we check it. The seventh stage is guidance on bibliometric analysis with maps from VOSviewer or Biblioshiny.
- How long does a systematic review take?
It runs into months. The protocol by itself needs 2–3 weeks. Title and abstract screening needs 3–8 weeks, and synthesis with the write-up takes another 4–8 weeks. Registry processing goes on whatever schedule the registry follows.
- Do you do the review?
No. You run the review. The question is yours, and so are the searches, the screening, the eligibility judgements and the interpretation. We review and check each step. Your methods section has to say honestly who did which part, and our ethics policy has the rest.
- Will I definitely get a meta-analysis?
That depends on whether the included studies are similar enough to pool. Suppose the populations or the outcome measures don’t match. Then we tell you that clearly, and you build a narrative synthesis instead, with our feedback.
- How is it priced and started?
The first step is a free consultation where we discuss about your question and any draft protocol you have. Then we give in writing a fixed price and revision rounds for each stage you pick. Sometimes the record count cannot be known yet. In that case we quote for protocol and searching first, and screening gets quoted once the real numbers come in.
Scope
What’s included
Each stage can be taken separately, and most scholars need only a few of them. You do the work at every stage, because the review belongs to you. Our part is review, training and checking.
Protocol review
You frame a focused question with PICO, PEO, SPIDER or some similar framework. After that you write the eligibility criteria and planned methods. We review the protocol against PRISMA-P, so it is ready for you to register on PROSPERO or OSF Registries.
Search string review
You build separate search strings for each database, using controlled vocabulary (MeSH, Emtree) along with free-text terms. We check them and show you where a string is missing terms. Your search log has to show dates and hit counts, and we check that also. Grey literature and citation chasing we help you plan too.
Screening training and checks
We show you how to remove duplicates and set up the Rayyan or Covidence project. Then we train you and your second screener in a pilot round to calibrate the criteria. The two independent screeners, both from your side, then follow a documented process. We check how conflicts were resolved and how far the two of you agree.
Extraction and appraisal checks
You pilot the extraction form first, and we review it. Appraisal is done with whichever tool suits your study designs. For randomised trials that means RoB 2, for non-randomised studies ROBINS-I, and for qualitative and other designs the CASP and JBI checklists. We check a sample of your extraction and appraisal judgements against the papers.
Help with synthesis and meta-analysis
You arrange a narrative synthesis by outcome or by theme. When the studies are comparable, you run a meta-analysis instead, with heterogeneity assessment and sensitivity analyses. We guide you through the R or RevMan steps and check your files and output, so you can run everything again yourself.
PRISMA reporting review
Your manuscript or chapter is checked against the full PRISMA 2020 checklist. We check that the numbers in your flow diagram add up. You fill the checklist with page numbers and we check it, to use when you submit to the journal.
Bibliometric analysis guidance
You run a Scopus or Web of Science search, record it and export the full records. We guide you as you merge and deduplicate these records and make a thesaurus file to clean up the different forms of author names and keywords. Only then you build the maps. In VOSviewer that means co-authorship, co-citation, coupling or keyword maps, and in Biblioshiny it means performance statistics and thematic maps. We check the maps and your notes on the central papers in each cluster, so the interpretation you write is one you can explain and defend.
Quality
How quality is verified
These are the checks every piece of work goes through before it reaches you.
Search strings re-run before reporting
We run every search string you wrote again in each database. We confirm the hit counts in your log and see that the final search date is noted down. The PRISMA mismatch that comes up most commonly is a count which nobody can reproduce later.
Flow diagram numbers reconciled
Each box of the PRISMA flow diagram is totalled and checked against the screening export. For every full-text exclusion, the reason is recorded.
Screening agreement measured
Once the pilot round is over, we calculate how far the screeners agree, for example using Cohen’s kappa. You refine the criteria at that point itself, before full screening starts.
Right appraisal tool for each design
We check that your risk-of-bias tool suits the study design, and that every judgement is written along with the supporting text, so you can trace each assessment back to the paper it came from.
Meta-analysis only when justified
Before any pooling, we check with you whether the populations, interventions and outcome measures can really be compared. If they cannot, we tell you so and you go for a narrative synthesis.
Checklist checked item by item
We go through your completed PRISMA 2020 checklist and check that each item appears in your text where the checklist says. Both you and the journal can check compliance from it.
Time
Realistic timelines
Keep months in your plan. Studies that looked at published systematic reviews in health research found they take well over a year on average, counting from registration till publication. A review for a thesis chapter with a narrow question can finish faster. Even then it usually goes on for some months.
| Stage | Typical time | What it depends on |
|---|---|---|
| Your protocol, and our review | 2–3 weeks | How settled the question is, and supervisor sign-off |
| Registration (PROSPERO, where eligible) | Several weeks, sometimes longer | The registry’s own processing, which is outside our control |
| Search strings, our check and your searches | 2–4 weeks | Number of databases and complexity of the topic |
| Title and abstract screening | 3–8 weeks | Number of records (often 2,000 to 10,000) |
| Full-text screening | 2–6 weeks | Number of full texts and how easily they can be obtained |
| Extraction and risk of bias | 3–8 weeks | Number of included studies and complexity of outcomes |
| Synthesis, meta-analysis and write-up | 4–8 weeks | Whether pooling is possible, and revision rounds with you |
Before you search
Registration: PROSPERO, OSF or neither?
Many journals expect registration, and it is good practice also. But each registry accepts only certain kinds of review.
PROSPERO registers systematic reviews which have an outcome directly relevant to human health. Many reviews in medicine, nursing, public health, psychology and allied fields come under this. But PROSPERO will not take a scoping review. A literature review which is not systematic also gets refused there. PROSPERO also expects you to register before screening starts, so a review that has already finished extraction may get refused. Reviews in management, education, engineering or computer science usually fall outside its scope.
For such reviews you can go to OSF Registries. It takes pre-registered protocols from any discipline without any charge, and you get a time-stamped public record. Another option is to publish the protocol as a paper in a protocol journal. Either way, you are showing that the question and methods were fixed before you saw any results.
Systematic review, scoping review or narrative review?
Some scholars have a question about what research exists on a broad topic. Their question is not really about what the evidence shows for one specific effect. For them a scoping review (reported to PRISMA-ScR) may be the correct label. Some departments only need a thorough, well-argued thesis chapter. For them a narrative review with good documentation is usually enough, and our literature review structuring and feedback service covers that. In the consultation itself we tell you the type, because editors look at this label when screening.
When a meta-analysis makes sense
A meta-analysis needs at least a few studies that measure comparable outcomes in comparable populations. The effect sizes and variances should be reported, or obtainable. Then you run it, with random-effects or fixed-effect model as per the data, and we check the model choice. Heterogeneity has to be reported (such as I²). Where the number of studies is enough, small-study effects are checked too. If the conditions are not met, you go for a narrative synthesis. Heavier statistics go under our statistical analysis service.
Pricing
How pricing works
The size of a systematic review can vary a lot. One narrow question may give 800 records and 12 included studies. A broad one may give 9,000 records and 80 studies. That is why we quote only after seeing your question, a draft protocol if you have one, and a rough idea of how much evidence is out there. You get the quote in writing before we begin. After that it is fixed for the stages you choose, it does not go up midway.
Sometimes nobody can say in advance how big the screening will be. Then the protocol and search stages are quoted first, and screening after the real record count is in front of both of us. That way the price matches the work that is actually there.
What the quote depends on
- The stages you need
- How many records come back and how many studies get included
- How many databases, and how complex the search strings are
- The study designs involved and the appraisal tools they need
- Whether a meta-analysis is possible and needed
Every quote includes
- A written scope listing exactly what will be delivered
- A fixed price agreed before any payment
- Payment in stages, tied to delivery
- Revision rounds stated in the scope
Process
How it works
Free consultation
A call or message with a specialist in your subject. We review your material and say plainly whether we are the right fit.
Written scope and quote
You get a written scope, a timeline and a fixed price. Work and payment begin once you agree to it.
Staged delivery
Work arrives in stages you review, each with a similarity report where text is involved. You pay stage by stage.
Revisions and hand-over
Revision rounds are included. At the end you get the final files, and an explanation of anything you’ll need to defend.
Keep reading
Guides and related services
How to do a systematic review: the PRISMA 2020 steps
The PRISMA 2020 steps from protocol to flow diagram, for a PhD chapter or a standalone paper.
How to write a literature review
How to move from summarising papers to building an argument that ends at your research gap.
Literature Review Structuring and Feedback
A review that synthesises and argues toward your gap, rather than summarising paper after paper.
Writing a research gap statement
The kinds of gap, a sentence pattern that works, and weak-versus-strong examples.
FAQ
Frequently asked questions
Can a systematic review be done in a month?
Mostly no. Protocol, searching, dual screening, appraisal and synthesis each need some weeks, and PRISMA wants every one of them reported. A rapid review with a tight scope can be done in less time. But then it should be labelled as a rapid review only.
Do I have to register on PROSPERO?
Only if your review is eligible and your journal or university expects it. Many health journals do expect it. When the topic is outside the health remit of PROSPERO, OSF Registries is a common alternative. Keep registration in your plan from the first month, because it has to be done before screening begins.
I only found six studies. Can I still do a meta-analysis?
Possibly, if those six studies measure comparable outcomes in comparable ways. With six only, funnel plots for publication bias are generally not meaningful. Many times a carefully arranged narrative synthesis is easier to defend. We will go through the studies and advise which approach fits.
Is a bibliometric analysis the same as a systematic review?
No, they are different methods. A bibliometric analysis maps publication records. It shows who writes on a topic, in which journals, whom they cite and which keywords appear together. In a systematic review you start from one focused question. To answer it you appraise the findings of each study you included. Some papers combine both, and the maps guide a close reading of the central studies. Many bibliometric papers get rejected for one reason, maps is shown but nobody interprets them. So the reading part you cannot skip. For the method itself you can read our bibliometric analysis guide. In the consultation we will tell you which approach your question needs.
Rayyan or Covidence?
Both of them support blinded dual screening. Rayyan has a free tier and many students use it. Covidence works on subscription, but lots of universities have a licence, and it handles extraction and risk-of-bias forms too. Mostly we say take the one your university already pays for.
What if the examiner asks why a study was excluded?
If your screening record is proper, you can give an exact answer. A reason is logged against every full-text exclusion. We train you to set up the process so that this record exists and matches with the flow diagram in your thesis.
Will my review idea be kept confidential?
Yes. Your question, protocol and data are shared only with the people working on your review. We use them for your project only and delete them afterwards if you ask. If you want an NDA, we will sign one.
Talk it through with a specialist for free
Send us where you are and what you need. We’ll reply within one working day with clear advice, a written scope and a fixed quote. No obligation.
Book a free consultation