Home/Insights/Evidence synthesis
Evidence synthesis standards
26 reference pages on evidence synthesis. The methods manuals a review is held to, the reporting guideline for each review type, the tools for appraisal and extraction, and the biases that survive synthesis.
Every named resource here — the Cochrane Handbook, the IOM standards, the AHRQ Methods Guide, MOOSE, JBI, EQUATOR, PRISMA 2020, SRDR+ — was checked against its issuing body rather than described from memory.
Methods manuals and handbooks 5
Cochrane Handbook for Systematic Reviews of InterventionsThe reference methods manual for intervention reviews, and the source of the tools most journals now expect a review to have used.Finding What Works in Health Care: Standards for Systematic ReviewsThe US Institute of Medicine’s standards for how a systematic review should be conducted — governance and process, not just method.AHRQ Methods Guide for Effectiveness and Comparative Effectiveness ReviewsThe methods manual behind AHRQ’s Evidence-based Practice Center reviews, and the most detailed public guidance on comparative effectiveness questions.AAFP Clinical Practice Guideline ManualHow the American Academy of Family Physicians develops and endorses guidelines — and why it declines to endorse some.JBI Manual for Evidence Synthesis: scoping reviewsThe methodological source for scoping reviews, and the clearest statement of when a scoping review is the wrong choice.
Reporting guidelines 5
EQUATOR Network: enhancing the quality and transparency of health researchThe library where reporting guidelines live. Finding the right checklist before you write is the cheapest improvement available to a manuscript.EQUATOR guidelines for reviews, meta-analyses, overviews and HTAWhich checklist applies to which kind of review — the distinction reviewers test first, because the labels are used loosely.MOOSE: meta-analyses of observational studies in epidemiologyThe reporting checklist for meta-analysing observational studies — where confounding does not average out, however many studies you pool.Standards for reporting systematic reviews: PRISMA 2020The reporting standard most journals now require, substantially revised in 2020 — and a 2009-era review written to the old checklist will look dated.STROBE, MOOSE, PRISMA and CONSORT: what each one governsFour checklists, four designs. Foot & Ankle International is one of many journals that now require the appropriate one at submission.
Bias and critical appraisal 3
Evaluating risk of bias in systematic reviews of interventionsRisk of bias is judged per outcome, with a named tool matched to the design. A generic "quality score" is not an assessment.Publication bias in a systematic reviewStudies with positive results are more likely to be published, published sooner and published in English — so a review of the published literature over-estimates effects by construction.Assessing publication bias in clinical researchThe problem begins before the review does: trials that are conducted and never reported make the published record an incomplete account of what was tried.
Synthesis and analysis 4
What are the different types of reviews?Narrative, scoping, rapid, umbrella, systematic — the names are used loosely, and the one you claim sets the standard you are judged against.What is a meta-analysis in Cochrane reviews?Pooling results into a single estimate — and the harder editorial judgement is whether the studies should be combined at all.Quantitative synthesis: an updateThe AHRQ methods chapter on when and how to combine results — including the circumstances in which the right answer is not to.Choosing the data analysis for a research paperThe analysis follows from the design, the outcome type and the structure of the data — not from what is familiar.
Tools and data repositories 5
OpenMeta-Analyst: open-source meta-analysis softwareA free, cross-platform meta-analysis package from Brown University — useful, and worth checking the current maintenance status before adopting it for a long project.Systematic Review Data Repository Plus (SRDR+)A free AHRQ-funded platform for extracting, storing and sharing systematic review data — so the extraction behind a review does not die with the project.A standardised library of depression outcome measuresWhen every trial measures depression with a different instrument, the results cannot be compared — which is the problem standardised measure libraries and core outcome sets exist to solve.Integrating health system data with systematic reviewsReviews tell you what worked in trials; health system data tell you what is happening in your population. Each answers what the other cannot.Machine learning for screening in systematic reviewsAutomation can cut screening workload substantially. The question is not whether it is accurate but what recall you are willing to accept, because a missed study is invisible.
Writing, peer review and data collection 4
The rules for manuscript peer reviewersConfidentiality, competing interests, and a review that could be shown to the author without embarrassment.Professional medical writing supportProfessional writers are permitted and are associated with better reporting. What is not permitted is concealing them.English editing guidelines for authorsLanguage editing should make the argument legible without changing what it claims — and the line between those two is where most of the judgement lies.What medical data is collected through survey researchSurveys capture what people report — symptoms, behaviour, experience, quality of life. The design question is whether what they report is what you need to know.