How Do You Find Unpublished or Ongoing Studies?

How do you find research that has not been published?

Searching bibliographic databases tells you what has entered the literature those databases index. It does not necessarily tell you everything that has been studied.

A clinical trial may be registered and recruiting. Another may be completed but have no journal article. Researchers may have presented findings at a conference without subsequently publishing them in full. A dissertation may contain an otherwise unpublished study. A research team may have completed a project whose results remain available only from a registry, institutional report, sponsor, regulatory source, or the investigators themselves.

This creates an unusual search problem. You are trying to find research that may not have the title, abstract, keywords, citation record, or bibliographic metadata that normally make published studies discoverable.

There is no single database of everything researchers have started or completed. Finding unpublished and ongoing studies therefore requires you to search for evidence that a study exists, not merely for publications reporting its results.

Search study records and research trails, not only publications

Look for the study before looking for the paper

Unpublished and ongoing studies are not the same thing

These categories overlap in search strategies, but they describe different situations.

A third category complicates matters: completed studies whose results are available somewhere other than a journal article. A trial registry may contain posted results, for example. A dissertation or technical report may fully describe the study. Calling such research simply "unpublished" can obscure the fact that useful information is publicly accessible.

For a review, the more practical questions are: Does the study exist? What is its status? What reports or results are available? Does it meet the eligibility criteria?

Why look for studies that are not in journals?

The methodological reason is that publication is not a neutral filter.

Research may remain unpublished for many reasons, and the probability or timing of publication can sometimes be associated with study findings. A review restricted to readily published evidence may therefore provide an incomplete or systematically distorted representation of the research that was actually conducted.

This is part of the broader problem of publication bias in the published literature. Searching beyond journal articles cannot guarantee that every missing study will be recovered, but it can expose research that a conventional database search would never identify.

Ongoing studies matter for another reason. Even though they cannot usually provide final results yet, they show what evidence is coming. This can help researchers judge whether a review is likely to become outdated quickly, whether a research gap is genuinely unstudied, or whether several relevant studies are already underway.

Trial and study registries are among the most important starting points

For clinical and health-related research, trial registries are central sources for identifying planned, ongoing, completed, terminated, and sometimes unpublished studies.

Cochrane requires searches of relevant trial registers and results repositories for intervention reviews where applicable. Its guidance identifies ClinicalTrials.gov and the World Health Organization International Clinical Trials Registry Platform (WHO ICTRP) as important sources, alongside other topic-appropriate registers.

Registries can contain structured information such as:

• study title;
• registration number;
• sponsor;
• investigators or contacts;
• study design;
• eligibility criteria;
• interventions and comparison groups;
• planned outcomes;
• recruitment status;
• start and completion dates;
• study locations;
• links to publications;
• posted results for some studies.

That information can reveal a study even when no journal article exists.

ClinicalTrials.gov and WHO ICTRP are related but should not be treated as interchangeable

WHO ICTRP provides a search portal across data supplied by multiple registries. ClinicalTrials.gov is one of the major individual registries represented in the broader registry ecosystem.

Cochrane's technical guidance recommends searching ClinicalTrials.gov and WHO ICTRP separately for intervention reviews rather than assuming that searching one interface completely substitutes for the other. Differences in functionality and records can matter during retrieval.

Depending on the topic and jurisdiction, national or discipline-specific registries may also be relevant.

The principle is similar to bibliographic searching: overlapping sources do not necessarily provide identical retrieval.

Registry searching requires a different search strategy from MEDLINE or Scopus

Study registry interfaces are often less sophisticated than major bibliographic databases. Long Boolean strategies copied directly from MEDLINE may perform poorly.

Guidance on systematic registry searching recommends constructing searches around the essential concepts and adapting them to the capabilities of each register. Multiple simpler searches may retrieve studies that one highly restrictive query misses.

Consider variations in:

• intervention names;
• generic and brand names;
• condition terminology;
• acronyms;
• older terminology;
• spelling variants;
• study identifiers.

Search sensitivity is often more valuable than elegance. Registries were primarily developed to register research, not to make systematic reviewers happy with their Boolean syntax.

Do not rely completely on the status shown in a registry

A registry may label a study as recruiting, active, completed, terminated, withdrawn, or unknown. These labels are useful, but they are not infallible.

Cochrane notes empirical evidence that registry records are sometimes inaccurate or not kept up to date, including inaccurate completion status. A study labelled "recruiting" may have finished long ago if the record was never updated.

If the study is important, corroborate its status using:

• dates in the registry record;
• linked publications;
• recent registry updates;
• researcher or institutional webpages;
• conference records;
• author contact.

Treat registry status as evidence, not omniscience.

A completed registry record does not necessarily mean results are available

This distinction is fundamental.

A registry may establish that a study was completed without providing any numerical results. In other cases, structured results may have been posted even though no journal article exists.

Identification still has value when results cannot be obtained. It can reveal how much evidence remains unavailable and can therefore inform judgments about possible missing-evidence bias.

Use registration numbers as study fingerprints

Once you identify a registration number, search it broadly.

Registration identifiers often survive title changes and can connect otherwise difficult-to-match documents. Search the identifier in:

• bibliographic databases;
• Google Scholar;
• general web search engines;
• conference programs;
• preprint servers;
• institutional repositories;
• publisher websites.

A registration number can lead from an apparently unpublished registry entry to a journal article with a completely different title.

It can also connect a conference abstract, preprint, protocol, results paper, secondary analysis, and follow-up publication to the same underlying study.

Conference records can reveal completed research before or without full publication

Conference abstracts and proceedings are another important source of less visible studies. Researchers may present findings years before full publication, and some presentations never progress to journal articles.

Search:

• conference abstract databases where available;
• conference proceedings;
• professional society websites;
• conference program archives;
• supplements of journals that publish conference abstracts;
• authors and study identifiers found in conference records.

A conference abstract may provide enough information to identify a study even when it does not provide enough detail for inclusion in a synthesis. If no later report appears, the process for tracing an abstract without a full publication becomes relevant.

Theses, dissertations, reports, and working papers may contain otherwise unpublished studies

Research that never became a journal article may still have been documented extensively elsewhere.

A doctoral dissertation can contain several complete studies. A government or institutional report may describe a program evaluation. A working paper may circulate findings that never proceed to journal publication. A technical report may contain methods and results unavailable in any conventional article.

Search institutional repositories, dissertation collections, organizational websites, disciplinary repositories, and targeted web sources according to the type of evidence expected. The methods for finding these forms of grey literature complement registry searching because many studies are not formally registered.

Preprints occupy a useful middle ground

A preprint is not hidden research. It is publicly available. But it can reveal research before conventional journal publication and may therefore help identify recently completed studies.

For rapidly developing topics, preprint searching can reduce the lag between study completion and discoverability. If preprints are eligible for your review, handle their version status explicitly and check whether a journal article has subsequently appeared.

The decision about whether to include preprints in the literature review should remain separate from their usefulness for discovering studies.

Regulatory sources can contain studies or results absent from journals

For reviews of drugs, medical devices, and other regulated health interventions, regulatory agency sources can provide evidence not available in conventional publications.

Cochrane identifies regulatory agency sources and clinical study reports as potentially valuable sources of unpublished information. Depending on the jurisdiction and product, regulators may provide assessment documents, study information, results, or clinical study reports.

These sources can be particularly important when journal publications provide only a subset of the evidence submitted during regulatory evaluation.

Access, organization, and searchability vary considerably by regulator and product, so this approach is more specialized than ordinary registry searching.

Search the websites of likely sponsors and research organizations

Think about who would have commissioned, funded, registered, or conducted the study.

Potential sources include:

• government agencies;
• universities;
• research institutes;
• charities and foundations;
• professional organizations;
• commercial sponsors;
• international organizations;
• research consortia.

Search their project pages, publications, research reports, trial disclosures, repositories, and archived materials.

This is especially useful when a known project or funding program appears to have generated research but no conventional publication can be located.

Funding databases can reveal projects before publications appear

In some fields and countries, research funders maintain searchable databases of funded projects. These may provide project titles, investigators, institutions, abstracts, award periods, and funding amounts.

A funding record does not prove that a study was completed or that results exist. It can, however, reveal research activity and give you names, institutions, project identifiers, and dates to pursue elsewhere.

This can be valuable when investigating an emerging field where funded work is underway but journal publication lags behind.

Reference lists can reveal research that databases do not

Published articles and reviews sometimes mention unpublished studies, personal communications, conference presentations, dissertations, reports, trial identifiers, or "data on file."

When you encounter such a reference, treat it as a lead.

Search:

• the cited title or description;
• the authors;
• the institution;
• the sponsor;
• the study acronym;
• the registration number;
• distinctive intervention or population details.

Older systematic reviews can be particularly useful maps because their authors may have contacted investigators when the study was easier to trace.

Contacting study investigators can identify research no database contains

Some unpublished studies are discoverable only through people.

Campbell systematic-review guidance notes that colleagues and informal communication can sometimes be the only means of identifying unpublished data. It suggests contacting authors of relevant studies and experts in the field to ask about additional published or unpublished research.

Contact can be particularly useful when:

• a registry shows a completed study but no results;
• a conference abstract has no full publication;
• a published paper mentions an unpublished companion study;
• a research program appears to have produced additional work;
• eligibility cannot be established from the available record;
• results needed for synthesis are missing.

Make the request specific. Identify the study or research question and explain what information you need.

Experts can help identify studies, but expert knowledge is not a reproducible database

Researchers active in a field may know about ongoing projects, abandoned studies, dissertations, conference presentations, or research groups that are difficult to discover electronically.

This can be useful, particularly in specialized fields. But personal networks are uneven. One expert's knowledge is not necessarily representative of the whole field.

For systematic reviews, document who was contacted, when, and what was requested. Expert contact should supplement structured searching rather than become an undocumented substitute for it.

Commercial sponsors can be relevant sources

When companies sponsor research, they may possess records or results not represented in journal publications. Depending on the topic, company trial registries, disclosure portals, clinical study reports, regulatory submissions, or direct contact may provide additional evidence.

This is particularly relevant in pharmaceutical and medical-device research, where reporting bias has historically been a major methodological concern.

Do not assume either that sponsor-provided data are invalid because they come from industry or that they are reliable because they are more detailed than journal reports. Provenance, methods, completeness, and potential conflicts still require appraisal.

Search by study identity once you find a lead

A productive search often changes character after the first clue.

Suppose a registry reveals an unpublished trial. Stop searching only by your broad review keywords and start tracing that specific study.

Use combinations of:

• registration number;
• principal investigator;
• study acronym;
• sponsor;
• institution;
• intervention;
• location;
• recruitment dates;
• sample size.

This can uncover publications whose titles bear little resemblance to the registry title.

Do not confuse multiple reports with multiple studies

Tracing unpublished research often produces several documents for the same project: registry record, conference abstract, preprint, dissertation, report, and eventual journal article.

Link them before counting studies.

Cochrane emphasizes that the study, rather than the individual report, should remain the unit of interest. Otherwise, the same participants can accidentally enter a synthesis more than once.

Identifiers such as registration numbers are particularly valuable for avoiding this problem.

An ongoing study should not be treated as though it already has findings

Knowing that a study exists does not permit you to infer its results.

If a registry says a randomized trial is recruiting, you can report that relevant research is underway. You cannot treat the intervention as supported or unsupported by that study.

Similarly, a completed study without available results is evidence about the existence of missing evidence, not evidence about the direction of the missing findings.

Unpublished results may be too incomplete to use

Even when you identify a study, the available information may not support inclusion in a meta-analysis or detailed synthesis.

A registry might provide the sample size but no outcome results. A conference abstract may report a p-value without an effect estimate. An investigator may confirm that a study was completed but no longer have accessible data.

Do not manufacture missing information.

The study may still matter when assessing the completeness of the evidence base, but study identification and usable data retrieval are separate achievements.

Record what happened to every potentially relevant study

For a systematic review, unpublished and ongoing studies should not disappear into your search notes.

Record enough information to track:

• where the study was identified;
• its identifier;
• current status;
• reports located;
• whether results are available;
• eligibility status;
• author or sponsor contact attempts;
• whether the study was linked to a later publication.

This prevents repeated detective work and helps explain how the final evidence base was assembled.

Recheck ongoing and unresolved studies when the review is updated

An ongoing trial can become completed. A completed trial can post results. A conference abstract can become a journal article. A preprint can be peer reviewed and published.

For reviews conducted over several months or updated before publication, revisit important unresolved records.

Trial-register guidance also emphasizes updated searches because identifying forthcoming evidence can inform the timing of traditional or living systematic-review updates.

The literature does not politely freeze while your manuscript is with the co-authors.

There is no guaranteed complete search for unpublished research

This limitation should remain explicit.

Campbell guidance notes that there is no easy and reliable way to obtain information about every study that was completed but never published. Registration practices differ across fields. Older studies may predate registration requirements. Conference archives can disappear. Investigators may not respond. Organizations may not retain public records.

A thorough search can therefore reduce uncertainty without eliminating it.

This is why the question of when searching for unpublished research is worth the effort matters. Search intensity should reflect the purpose of the review, the plausible risk of missing evidence, and the resources available.

How to trace a study that never appears in the journal search

Common mistakes when searching for unpublished and ongoing research

Build a search around where evidence of the study could exist

The appropriate search is driven by your field and review question. Clinical intervention research may have strong registration infrastructure. Other disciplines may rely much more heavily on repositories, conference records, funding databases, research networks, and direct investigator contact.

For a systematic review, decide these approaches prospectively where possible and document them. Searching for unpublished research can become open-ended, so the protocol should identify the sources and methods proportionate to the risk that missing evidence could affect the review.

And keep one conceptual boundary firm: finding evidence that a study exists is valuable, but it does not give you permission to invent the results you wish the investigators had reported.

Before concluding that you have looked for unpublished and ongoing studies

Frequently asked questions about finding unpublished and ongoing studies

Search for traces of the study, not only for papers

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.