Skill reference: systematic-review

Skill reference: systematic-review

The systematic-review skill authors one document genre: a PRISMA 2020 systematic review — a reproducible, auditable account of how a body of evidence was identified, screened, appraised, and synthesised, centered on a mandatory Mermaid PRISMA flow diagram that reconciles record counts at every stage. This reference describes what that document type is, how the skill produces one, when it earns its place, and the provenance behind it.

Property	Value
Authors	A PRISMA 2020 systematic review
Purpose group	Scholarly writing
MIF conceptType	semantic
Target MIF level	3
Primary source	The PRISMA 2020 statement, Page et al., BMJ 2021

What this document type is

A systematic review, in the PRISMA 2020 sense, is a structured account of a review question answered by exhaustively identifying, screening, appraising, and synthesising the available evidence — not a narrative summary of whatever studies the author happened to read. Its defining trait is the PRISMA flow diagram: a reconciling count of records identified, screened, excluded (with reasons), and studies included, without which the report is not conformant. The review also carries a structured abstract, a reproducible Methods section (eligibility criteria, information sources, search strategy, selection process, data items, and a named risk-of-bias tool), a risk-of-bias table alongside the flow diagram, and an explicit Registration & Protocol statement even when none exists.

This is distinct from a general scholarly write-up with no mandatory flow diagram or registered search protocol (an academic report), from a single already-made technical decision (an adr), and from a requirements document scoping what to build (a prd or feature-spec). It is reasoned synthesis of existing evidence, so it projects to MIF as semantic content at Level 3.

How the skill produces one

systematic-review is a genre skill: it carries the PRISMA 2020 pattern as durable instructions plus exemplars, and writes the artifact over a MIF floor so the result is at once a human-readable review and a machine-conformant unit.

• Pattern, made operational. The skill encodes the structured abstract, the reproducible Methods subsections, and treats the Mermaid flowchart TD PRISMA diagram as mandatory and reconciling — records identified minus removals at each stage must equal studies included, or the diagram is a defect. It directs authors to verify the current PRISMA guidance live before writing, rather than trusting a fixed item count.
• Exemplars set the bar. Like every genre in the suite it ships good-l1.md (the MIF Level-1 floor), good.md (the Level-3 target), bad.md (a counter-example that reports a study count with no flow diagram), and evals/evals.json. The check-exemplars gate proves good-l1.md validates at L1 and good.md at Level 3.
• MIF projection. The document is authored with MIF frontmatter (via the shared mif-frontmatter substrate) and a conceptType of semantic. mif-validate proves the Markdown ↔ JSON-LD round-trip is lossless before the document is considered done.

When it is beneficial

Reach for systematic-review when the deliverable must make the evidence-selection process legible and reproducible end to end — another reviewer must be able to repeat the search and selection from the Methods prose alone, and every included or excluded record must be traceable through the flow diagram. It earns its place where the review’s credibility rests on its process, not just its conclusions: clinical, policy, or engineering questions where cherry-picked evidence would be a defect.

Do not use it for a general scholarly write-up with no mandatory flow diagram or registered search protocol — that is academic: a formal research report with a selectable citation style but no PRISMA-mandated flow diagram or reproducible multi-stage search protocol. Do not use it for a single already-made technical decision — that is an adr. Do not use it to scope what to build and why before design — that is a prd or feature-spec. If the alignment is a trade-off narrative around one proposed design rather than a synthesis of existing evidence, an engineering report or google-design-doc fits better.

Example

A review titled “Telehealth-Delivered Cognitive Behavioral Therapy for Adult Depression” poses a PICO question — does telehealth CBT produce comparable depression-symptom outcomes to in-person CBT in adults with major depressive disorder — and searches MEDLINE, PsycINFO, and CENTRAL through 2026-05-01. The PRISMA flow diagram reconciles 1,842 records identified down to 12 included randomized controlled trials, with every exclusion at the full-text stage tagged with a reason (wrong population, intervention, outcome, or design). A Cochrane RoB 2 table appraises each included trial across five bias domains, and the synthesis reports that telehealth CBT was broadly equivalent to in-person CBT and superior to waitlist/care-as-usual, flagging the one “High”-risk-of-bias trial as a caution rather than folding it into the headline finding. The review states plainly that it was not prospectively registered, per PRISMA’s “Other information” reporting requirement.

Provenance & citations

• Genre source — PRISMA 2020: Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71. https://doi.org/10.1136/bmj.n71.
• Skill provenance: authored by the systematic-review skill in the mif-docs plugin, https://github.com/modeled-information-format/mif-docs-plugin; the skill’s exemplars and evals/ define and verify the pattern.
• MIF conformance: the document projects to canonical JSON-LD under the MIF specification, https://mif-spec.dev, and is proven lossless by mif-validate.
• Index: this skill is one entry in the skills by purpose catalog; its Scholarly writing siblings are documented alongside it as the genre-consolidation effort completes.