03 · What You Need to Know
Separate finding the study from using the abstract as evidence
A conference abstract is a report of research, not necessarily the complete research report
Conference abstracts are short summaries submitted for presentation at scientific or scholarly meetings. Depending on the conference, they may describe completed research, preliminary findings, ongoing studies, protocols, methodological work, reviews, case reports, or other scholarly activity.
They are commonly treated as part of grey-literature searching because they may sit outside conventional full-length journal publication. Some are subsequently published as full articles. Others never are.
The abstract is therefore best understood as a report of a study, not automatically as the study's definitive record. The same underlying research may eventually appear in several forms: a registry record, conference abstract, dissertation, preprint, working paper, and peer-reviewed article.
Identifying through a conference abstract
The abstract tells you that potentially eligible research exists and should be investigated further.
Including an abstract-only report
You use information reported in the abstract itself as part of the evidence included in the review.
This distinction resolves much of the apparent disagreement about conference abstracts. A reviewer may have strong methodological reasons to search conference records while remaining cautious about extracting results from abstracts that provide inadequate information.
Why search conference abstracts at all?
The main reason is that not all research presented at conferences subsequently becomes a full publication.
Research examining cohorts of conference abstracts has repeatedly found incomplete progression to full publication, although rates vary substantially among disciplines, conferences, study types, and observation periods. Some studies have also found that research with positive or statistically significant findings has a greater probability of full publication. A meta-analysis of empirical studies of publication bias reported that, among cohorts beginning with conference abstracts, studies with positive results had higher odds of subsequent publication than those without positive results.
This creates a potential problem for evidence synthesis. If reviewers search only full journal publications, research that remained at the conference stage can disappear from the visible evidence base. That is one reason Cochrane guidance for intervention reviews regards searching relevant conference abstract sources as highly desirable.
The methodological concern connects directly with the broader question of how excluding grey literature can bias an evidence synthesis. Conference searching may expose research that conventional journal databases do not adequately represent.
Why not simply include every relevant conference abstract?
Because identifying research and evaluating it require different amounts of information.
A typical abstract has severe space constraints. It may report the study objective, a brief description of methods, one or two outcomes, and a conclusion while omitting details needed to determine eligibility or appraise the research properly.
Potentially missing information includes:
- complete participant eligibility criteria;
- sampling and recruitment procedures;
- details of randomization, allocation, or blinding where relevant;
- complete descriptions of interventions or exposures;
- measurement instruments and outcome definitions;
- attrition and missing-data procedures;
- complete numerical results;
- prespecified versus exploratory analyses;
- harms or secondary outcomes;
- funding, conflicts of interest, and study limitations.
A 2022 methodological study examining conference abstracts considered for a large Cochrane network meta-analysis illustrates the problem. Of 90 abstracts examined, only 13 could be classified as eligible from the available information, while 77 remained awaiting classification because key eligibility information was missing. Most abstracts also lacked sufficient information on many data-extraction items.
That does not establish that conference abstracts are intrinsically unreliable. It shows that poor or incomplete reporting can prevent reviewers from knowing enough to make the decisions required by the review.
Peer review at a conference is not necessarily equivalent to journal peer review
Conference selection procedures vary widely. Some conferences use detailed peer review, others use brief abstract screening, and some primarily assess relevance to the meeting. Acceptance may be for oral presentation, poster presentation, proceedings publication, or another format.
You should therefore avoid assuming either that conference acceptance guarantees rigorous peer review or that conference material underwent no review at all. When the review process matters, investigate the conference's procedures rather than inferring them from the word "conference."
More importantly, peer-review status does not solve the reporting problem. Even a carefully reviewed abstract may still be too short to support detailed critical appraisal.
The abstract may report preliminary rather than final results
Research presented at a conference can represent an earlier stage of a study. Recruitment may later increase, follow-up may continue, analyses may change, errors may be corrected, outcomes may be added or reported differently, and conclusions may be revised before full publication.
Empirical comparisons have documented discrepancies between conference abstracts and corresponding full publications. For example, a study of conference abstracts describing systematic reviews in pain found some form of discordance in 40% of evaluable abstract-publication pairs, including differences in the direction of effect in some cases. Other research has similarly documented changes in reported outcomes and conclusions between conference and subsequent journal reports.
The implication is not that every conference result is wrong. It is that an abstract may represent a snapshot rather than the final research record.
Risk-of-bias assessment can be especially difficult from an abstract
A reviewer may be tempted to interpret missing methodological details as evidence that the method was not performed. That inference is unsafe. An abstract may omit allocation concealment, blinding, missing-data procedures, or analytic details simply because there was no space to report them.
Recent methodological research comparing conference abstracts with later full reports of randomized trials found that abstract-stage risk-of-bias assessments were frequently unclear because information was missing, and assessments could change once the full text became available. The authors cautioned that existing risk-of-bias tools may not be well suited to abstract-only evidence.
Watch Out
Do not convert "not reported in the abstract" into "the researchers did not do it." Absence of reporting and evidence of a methodological flaw are different problems. Both may limit how confidently you can use the abstract, but they should not be conflated.
Search for a full report before treating the study as abstract-only
When you find a potentially eligible conference abstract, the next step should usually be to determine whether the study has subsequently appeared elsewhere.
Search using:
- the exact abstract title;
- the first and senior authors' names;
- distinctive keywords from the title;
- sample size, setting, intervention, or study acronym;
- trial or study registration numbers when reported;
- author profiles and institutional repositories;
- citation databases and scholarly search engines.
Do not rely only on an identical title. Titles often change between conference presentation and journal publication. Matching should consider authors, participants, study design, intervention or exposure, location, sample size, recruitment period, registration number, and other identifying characteristics.
This is part of the broader task of tracing research across grey-literature sources.
What if you find the full paper?
If a corresponding full report exists, it will usually provide the more complete basis for determining eligibility, extracting data, and assessing methodological limitations. However, do not automatically discard the abstract without comparison.
The abstract can reveal an earlier sample size, outcome, analysis, or conclusion that differs from the final paper. Where differences are consequential, the two reports should be treated as multiple reports of the same underlying study rather than independent studies.
Counting the conference abstract and journal article as two separate studies would duplicate participants and distort the synthesis.
What if the abstract never became a full paper?
This is the difficult case. Failure to locate a full publication does not necessarily mean the research was poor, rejected, or abandoned. A systematic review of investigators' reported reasons for nonpublication after conference presentation found that lack of time was a commonly reported explanation, alongside resource constraints, low priority, incomplete studies, and other factors.
If no full report can be located, you need to decide whether the abstract contains sufficient information for your review. Where appropriate, contacting the authors may provide clarification, additional results, or information about subsequent publication.
The decision process becomes more important when a conference abstract never becomes a full paper, because the reviewer must distinguish between evidence that exists but is incompletely reported and evidence that cannot be used reliably.
Inclusion depends partly on what you intend to do with the evidence
"Include" can mean several things in a literature review. An abstract might be:
- recorded as evidence that a study exists;
- listed among studies awaiting classification;
- described narratively but not used in quantitative synthesis;
- included in evidence mapping;
- used for selected descriptive characteristics;
- included in a meta-analysis when sufficiently detailed numerical data are available.
These uses impose different information requirements. An abstract may be sufficient to establish that researchers investigated a topic but insufficient to estimate an effect size or assess risk of bias.
For this reason, inclusion should not be framed as a simple yes-or-no rule detached from the purpose of the review.
Systematic reviews need a predefined policy
If you are conducting a systematic review, decide how conference abstracts will be handled before seeing their results whenever possible.
Your protocol can specify whether conference abstracts are eligible, whether they will be used only to identify studies, what minimum information is required for inclusion, whether investigators will be contacted, and how later full publications will be handled.
This protects against an awkward form of selective decision-making: accepting abstract-only evidence when its findings support the emerging conclusion but rejecting it when the findings are inconvenient. Methodological rules should not change according to what the abstract reports. The peer reviewers tend to notice that sort of thing, usually around Reviewer 2 o'clock.
