Yes, you can cite non-peer-reviewed sources in a medical literature review when you label them clearly and appraise their quality.
Medical reviews aim to map the best available evidence with a transparent method. That usually starts with peer-reviewed studies, yet many projects also need data from preprints, conference proceedings, trial registries, or official reports. You can include these sources, but you need tight screening, clear labeling, and a plan that avoids bias.
Quick Answer, Scope, And When It Makes Sense
You can cite material that has not gone through formal journal review when it fills a gap, adds timely context, or tracks emerging work. That includes preprints on new pathogens, device safety alerts, drug shortage notices, and high-quality government assessments. The bar is higher, so treat each item as provisional and show readers exactly how you judged it.
Source Types You May Use And The Guardrails
Not all unrefereed items carry the same weight. The table below lists common source types you might bring into a clinical review, a quick “use/no-use” signal, and the minimum steps that keep your synthesis reliable.
| Source Type | Use In Medical Reviews? | What You Must Do |
|---|---|---|
| Preprints | Yes, with care | Flag as preprint; check methods, sample size, and version/date; look for later journal publication |
| Conference Abstracts | Sometimes | Verify authors and trial identifiers; seek slide deck or poster; treat findings as preliminary |
| Government Or Agency Reports | Often | Confirm issuing body and methodology; record edition/version; note whether the report is guideline, audit, or surveillance |
| Clinical Trial Registries | Yes | Extract registry ID, status, outcomes, and protocol changes; pair with publications when available |
| Manufacturer Safety Communications | Sometimes | Record lot/device codes; confirm recall class or alert level; look for independent corroboration |
| Theses/Dissertations | Sometimes | Check university archive, committee names, and methods; inspect full text and data appendices |
| News Articles Or Press Releases | Rarely | Use only for context or dates; never as outcome evidence; trace to the primary study |
| Institutional White Papers | Sometimes | Evaluate authorship, funding, and method; see whether the findings appear in peer-reviewed form |
| Practice Guidelines | Yes | Use the current version; note sponsoring society and evidence grading scheme |
Why Reviewers Still Start With Journal Articles
Peer-reviewed studies provide the core because they follow reporting norms, present full methods, and tend to be easier to appraise with standard tools. That does not mean every journal paper is sound, only that appraisal checklists fit them well. When you add gray literature, you must adapt those same checks to the format at hand.
Using Non-Peer-Reviewed Sources In A Clinical Review — Rules That Keep You Safe
Use a protocol. State up front which unrefereed sources you will search, how you will appraise them, and where they will appear in your synthesis. Keep the language neutral and specific. Readers should see what you did and be able to repeat it.
Plan Your Searches
Search standard databases and add targeted routes for gray material: preprint servers, conference libraries, agency sites, and registries. Save queries and dates. Capture versions and web archive links when the page may change.
Set Clear Eligibility Rules
Define which outcomes, populations, and designs are allowed. Decide when a preprint or abstract is enough to include. When the item lacks core methods, keep it for background only. When a preprint later appears in a journal, update the record and treat the journal version as the primary source.
Label Everything
In your tables and narrative, tag each item as “preprint,” “conference abstract,” “registry entry,” or the exact report type. Add the date you accessed it. Use consistent wording across figures and text so a reader can scan the status at a glance.
Appraise With Fit-For-Purpose Tools
Risk-of-bias tools were designed for full studies, so adapt them: judge selection, measurement, confounding, and reporting bias using what the item reveals. For guidelines and reports, check method sections and evidence grading schemes. When detail is missing, mark the domain as “unclear” and explain why.
Where To Place These Sources In Your Synthesis
Keep the hierarchy visible. Make pooled estimates from completed studies with usable methods. Use conference material, registries, or agency memos to fill context gaps, map safety signals, or explain trial pipelines. When a key conclusion rests on an unrefereed item, say so in the sentence that states the finding.
How To Cite And Report Transparently
Follow a medical style that handles preprints, abstracts, and reports cleanly. Record DOIs or registry IDs when present. Add version numbers and access dates for web-only items. When journal policy asks you to flag preprints, mirror that flag inside your reference list and in-text labels.
Two Author-Facing Standards You Should Know
Many journals and review teams align with the ICMJE’s “Recommendations.” That document outlines reference expectations and includes guidance about preprints. You can read the full text on the official site: ICMJE Recommendations. For search and selection methods, the Cochrane Handbook sets practical rules on locating gray material and documenting the process; see the online chapter on searching: Cochrane Handbook, Chapter 4.
Common Pitfalls When Citing Unrefereed Material
Relying On Headlines Or Press Copy
Press pieces summarize; they do not supply methods or raw data. If a news item points to a study or alert, trace the primary item and cite that instead.
Mixing Evidence Types In One Estimate
Do not pool a preprint with a completed trial in the same meta-analysis unless you have a strong method plan. Even then, run a sensitivity test that drops provisional items and show both results.
Hiding Status Labels
If readers cannot tell which items lack journal review, the whole review looks shaky. Keep the labels where eyes land: in tables, forest plots, and the main text.
Workflow That Keeps You Honest
Build A Tracking Sheet
Create one row per record with fields for source type, peer-review status, version, registry ID, outcomes, appraisal notes, and inclusion status. Add a field that links a preprint to its later journal article so updates are seamless.
Screen In Duplicate
Have two people screen titles/abstracts and full texts. Resolve clashes with a third reviewer or a ruleset you wrote before screening. Keep reasons for exclusion so your PRISMA flow can show counts at each step.
Extract What The Item Can Truly Support
From a preprint or abstract, grab the variables needed for your question and note what is missing. If key data cannot be verified, keep the record out of quantitative synthesis and explain the gap in your narrative.
How To Phrase Findings That Use Gray Literature
Use clear, plain text that signals provisional status. Here are patterns that keep readers oriented without hedging:
- “One preprint with 1,024 participants reported lower readmission; methods included X and Y.”
- “Two conference abstracts noted device failures; both lacked follow-up data.”
- “A national report showed year-over-year case growth; surveillance methods are described in the appendix.”
Weighting And Sensitivity Checks
Give greater weight to completed studies with full reporting. When gray-only evidence points in a new direction, run leave-one-out tests and show how the main estimate shifts when you remove provisional items. Report those results in the same section as your headline effect.
Documentation You Should Include
Readers need to retrace your steps. Keep the following artifacts and cite them in methods or supplements: the full search strings, all dates searched, a list of sites for gray material, the screening form, and the risk-of-bias template you used for non-standard sources.
Dealing With Updates And Version Drift
Preprints change. Agency pages get refreshed. Use web capture tools and record the date you accessed each page. When a preprint becomes a journal article, update the citation, move it to the peer-reviewed set, and adjust tables and figures that used its data.
Second Table: Appraisal Checks For Gray Literature
Use this compact checklist to judge non-refereed items and to document the call you made. These checks work across preprints, abstracts, and reports.
| Check | What To Look For | Tools/Notes |
|---|---|---|
| Authorship & Conflicts | Full names, affiliations, funding; conflict statement | Note funding links; match names to registry entries |
| Study Design & Methods | Clear design, sampling, outcomes, time points | Adapt risk-of-bias domains to the format |
| Data Availability | Tables, raw data, or appendices you can inspect | Record what is missing; mark domains “unclear” when needed |
| Version & Date | Version number, posting date, latest update | Capture web snapshots; add access date |
| Peer-Review Status | Preprint flag, conference review, internal review | Label in text and tables; do not bury the status |
| Outcome Fit | Measures match your review question | Exclude from meta-analysis if mismatch blocks pooling |
| Consistency With Other Evidence | Direction and size align with related studies | Plan sensitivity tests that drop gray items |
| Traceability | DOI, registry ID, or stable URL | Link preprints to later journal versions |
How This Plays Out In Practice
A vaccine safety review might include peer-reviewed cohort studies for the main estimate, registry entries to list ongoing trials, and an agency safety bulletin to note a rare device error. The bulletin would carry a clear label, sit in a context paragraph, and never anchor the pooled risk figure. That balance keeps the picture complete without overselling weaker material.
Reporting Language You Can Reuse
Methods Section
“We searched MEDLINE, Embase, CENTRAL, trial registries, preprint servers, and agency sites from inception to 1 June 2025. We included peer-reviewed studies for quantitative synthesis. We used non-peer-reviewed sources for context and for tracking in-progress work. We flagged their status in all tables and figures.”
Results Section
“The search returned 4,212 records; 68 met inclusion criteria for quantitative synthesis. We retained 11 preprints and 6 conference abstracts for qualitative context only. Two preprints were later published; we updated records and reran sensitivity checks.”
Limitations Section
“Some data came from preliminary reports with limited methods. These sources informed context and did not drive pooled effects. Where uncertainty remained, we marked risk-of-bias domains as unclear and ran leave-one-out tests.”
Bottom Line For Medical Authors
You can cite material that lacks journal review inside a medical literature review when you keep status labels visible, appraise carefully, and show how those items affect (or do not affect) your conclusions. Start with peer-reviewed studies, add gray sources for context and completeness, and document every step so readers can follow the thread.