Should You Include Preprints in a Systematic Review?

Preprints -- research manuscripts posted publicly before formal peer review -- have become a genuinely significant part of the research landscape in many fields, and deciding whether to include them in a systematic review search is a real methodological choice with specific trade-offs worth understanding rather than defaulting to either inclusion or exclusion without consideration.

The case for including preprints

Excluding preprints entirely risks missing genuinely relevant, completed research simply because it hasn't yet cleared a peer review process that can take many months. In fast-moving fields, or during time-sensitive situations like an emerging public health concern, excluding preprints can mean your review misses the most current available evidence, potentially reproducing a form of currency-related bias parallel to language bias or database-coverage bias.

The case against including preprints

Peer review, whatever its acknowledged limitations, provides a quality floor -- a baseline check of methodology and reporting that preprints, by definition, have not yet passed. Including unreviewed research risks incorporating studies with undetected methodological flaws that peer review would likely have caught, and preprints are also more likely to be later substantially revised, retracted, or never formally published at all, compared to research that has already cleared peer review.

A middle-ground approach many reviews adopt

A common and reasonable compromise is including preprints in your search and screening, but conducting your own more careful, explicit quality appraisal given the absence of peer review, and running a sensitivity analysis excluding preprint-only studies to see how much your pooled estimate depends on this less-vetted evidence. This lets you benefit from more comprehensive, current coverage while being transparent about how much your conclusions rest on unreviewed research specifically.

Checking for a preprint's eventual publication status

Before finalizing your review, checking whether an included preprint has since been formally published, and whether the published version differs meaningfully from the preprint version you originally screened, is worth doing at the update-search stage. Using the more complete, peer-reviewed published version where one now exists, rather than the earlier preprint, is generally the more defensible choice once both are available.

Preprint servers relevant to systematic review searches

medRxiv and bioRxiv are the most commonly searched preprint servers for health and life sciences research, and SSRN serves a similar function for social science and some economics research. Whichever preprint servers are relevant to your specific field should be named explicitly in your search strategy if you've decided to include them, with the same reproducibility expectations applied to any other database in your search.

Reporting your preprint decision transparently

Whatever you decide, state your approach to preprints explicitly in your methods section -- whether you searched preprint servers, how many included studies were preprints at the time of your search, and whether you conducted any sensitivity analysis specifically examining their influence on your pooled result. This transparency lets readers judge for themselves how much weight to place on conclusions that may rest partly on not-yet-peer-reviewed evidence.

Risk-of-bias assessment for preprints specifically

Applying your standard risk-of-bias tool to a preprint works the same way as for any other included study, but it's worth being particularly careful and thorough given that the usual quality floor peer review provides hasn't yet been applied. Some reviewers apply a slightly more conservative or thorough appraisal specifically to preprint sources, given the absence of this external check, though this should be a genuine, careful appraisal rather than an automatic downgrade applied without actually examining the study's specific methodology.

Field-specific norms worth checking

Some fields and some target journals have explicit stated positions on including preprints in systematic reviews, and checking your specific field's current norms and your target journal's expectations before finalizing your approach can save a difficult conversation at the submission stage. This is an area where practice has evolved relatively quickly in recent years, making it worth verifying current expectations rather than relying on older training or a previous project's approach.

A reasonable default position

For most systematic reviews, searching preprint servers as part of a comprehensive search, applying careful risk-of-bias assessment to any included preprints, running a sensitivity analysis examining their specific influence, and transparently reporting all of this in your methods section represents a defensible, increasingly standard middle path -- neither ignoring a genuinely growing part of the research landscape nor treating unreviewed evidence with the same unquestioned confidence as peer-reviewed work.

Preprints and rapidly evolving topics specifically

For systematic reviews on genuinely fast-moving topics, where peer review's typical multi-month timeline means the published literature can lag meaningfully behind the actual current state of evidence, the case for including preprints, with appropriately careful appraisal, becomes considerably stronger than for a more slowly evolving research area where peer-reviewed evidence alone is likely to be reasonably current and complete. Making this judgment explicitly, rather than defaulting reflexively to whichever approach your team happens to be more familiar with, is what separates a deliberate methodological choice from an unexamined habit carried over from a previous, possibly quite different project. Revisiting this specific decision explicitly at the protocol stage of every new review, rather than defaulting silently to whatever approach your team used last time regardless of whether the new topic's specific research landscape actually calls for the same choice, keeps your methodology genuinely fit for purpose project by project. This kind of deliberate, project-specific reconsideration, rather than habitual repetition of a prior approach, is a small discipline that keeps your overall methodology genuinely responsive to what each new review actually requires. The preprint question, like several other methodological choices covered elsewhere, ultimately rewards a team willing to make a deliberate, disclosed decision suited to its specific project, rather than a fixed policy applied unreflectively regardless of context. That same principle of deliberate, context-specific choice runs through nearly every methodological decision a systematic review requires. This same principle of deliberate, context-specific choice, rather than habitual default, runs through nearly every methodological decision covered across this whole body of guidance. Consistency in applying it is what ultimately makes the difference in practice.