01 · The Question
Which Paper Should Represent a Study When Several Publications Exist?
You have linked four publications to the same underlying study. One is the earliest paper, another reports the prespecified primary outcome, a third contains longer follow-up, and the fourth gives more methodological detail. Which one should be the primary report?
There is no universal bibliographic rule that the earliest, longest, newest, most cited, or highest-profile paper must automatically be primary. In a systematic review, the designation is primarily an organizational and methodological decision about which report will serve as the main source for the study's results.
Cochrane guidance requires review authors to choose and justify the report used as a source for study results when multiple reports exist. It also explicitly warns against discarding the remaining reports, because they may contain valuable information unavailable in the chosen primary source.
03 · What You Need to Know
What “Primary Report” Actually Means
First Confirm That the Publications Belong to the Same Study
Choosing a primary report makes sense only after you have established that the publications are reports of one underlying investigation. Shared registration numbers, recruitment settings and dates, participant characteristics, interventions, study identifiers, and other study-level features can help establish that relationship.
If that linkage remains uncertain, first determine whether you are actually looking at multiple reports of the same study. Otherwise, you risk collapsing genuinely independent studies into one record.
Primary Report Is a Review-Level Designation
Authors do not always label one publication “the primary report.” A study may instead produce several journal articles emphasizing different outcomes or phases. The review team may therefore need to identify a main report for organizing the study record and sourcing results.
Cochrane's guidance is deliberately practical: multiple reports should be collated, and review authors should choose and justify which report will serve as the source for study results, particularly when reports contain conflicting results.
The word “primary” in this context should therefore not be confused with primary research versus secondary research. All of the linked publications may report primary research data. One is simply being designated as the main report within your review workflow.
Primary study report
The report designated as the main reference or principal source for study results within the review.
Secondary or companion report
Another report of the same study that may supply additional, complementary, follow-up, subgroup, methodological, or outcome information.
The Primary Results Publication Is Often a Strong Candidate
When a clearly identifiable full journal article reports the study's principal prespecified analysis for the main study population, it will often be a sensible candidate for the primary report.
Such a publication may contain the most complete account of participant flow, baseline characteristics, interventions or exposures, principal outcomes, and main analyses. It may also be the publication to which protocols, registries, and companion papers refer.
But “often” matters here. The most obvious article is not automatically the best source for every purpose, and systematic-review guidance does not establish a universal hierarchy in which one publication format always overrides every other source.
Completeness Matters, but Length Alone Does Not
A useful primary report should generally allow you to understand the study and its principal findings with relatively little reconstruction. Consider whether it clearly reports the population, design, intervention or exposure, comparator where relevant, follow-up, outcomes, participant flow, and results needed for your review.
A longer article is not necessarily more complete in the ways that matter. A lengthy secondary analysis can provide extensive theoretical discussion while omitting basic information about the original study. Conversely, a concise primary article may clearly document the core trial.
Relevance to Your Review Question Also Matters
Imagine a study with two detailed publications. The first focuses on a biological endpoint that is irrelevant to your review, while the second reports the educational or behavioral outcome your review is actually synthesizing.
It may be more practical to use the second report as the main source for the relevant result, even if another paper is historically regarded as the principal publication of the project.
This illustrates an important point: the report that best represents the overall study and the report that best supplies a particular data item need not always be the same.
Publication Date Is a Clue, Not a Rule
The earliest publication may be a protocol, conference abstract, preliminary analysis, or short report. A later full article may contain more complete results. Conversely, a later publication may simply report a subgroup or extended follow-up.
Do not designate the earliest or latest paper automatically. Instead, reconstruct the chronology and determine what each publication represents.
Journal Prestige and Citation Count Are Poor Selection Rules
A paper published in a more prestigious journal may be more visible, but journal status does not establish that it contains the most appropriate data for your review. Citation count is similarly unhelpful as a methodological criterion because citations measure attention rather than the suitability or completeness of a report for a particular synthesis.
The decision should be based on the relationship between the report, the underlying study, and the information required by your review.
Check Protocols and Registries When Available
A protocol or prospective registry entry can help you understand which outcomes and analyses were planned, how the study was designed, and whether a publication represents the principal results or a secondary analysis.
These sources can be particularly useful when several papers each appear substantial. They may also expose changes in outcome definitions, time points, or analytic populations that are difficult to recognize when publications are read separately.
One Report Does Not Need to Supply Every Data Item
Cochrane notes that a single source rarely contains complete information about a study and that multiple sources can sometimes contain inconsistent information. Data therefore often need to be collated and reconciled across reports.
Your study record might use one report for the principal outcome, another for adverse events, the protocol for planned methods, and a later follow-up publication for long-term results.
| Report Feature |
Why It May Support Primary Status |
Why It Is Not Decisive Alone |
| Reports principal study results |
Closely represents the main analysis of the underlying study |
Another source may report an outcome required by your review more completely |
| Full journal article |
Often provides more methodological and numerical detail |
Publication format does not guarantee completeness or accuracy |
| Largest or broadest study population |
May best represent the parent investigation |
A smaller analysis may be the relevant population for a particular review question |
| Most complete reporting |
Reduces reconstruction across sources |
No single report may be complete for every item |
| Earliest publication |
May represent the original results paper |
Could instead be preliminary or incomplete |
| Latest publication |
May contain longer follow-up or corrections |
Could be a secondary analysis rather than the main report |
| Most cited publication |
May be the best-known report |
Citation frequency does not determine methodological suitability |
Conflicting Reports Require More Than Choosing Your Favorite
Sometimes two reports provide different participant counts, effect estimates, outcome definitions, or numerical results. This is precisely where a transparent primary-report rule becomes important.
First determine whether the apparent discrepancy can be explained by different follow-up periods, analytic populations, outcome definitions, corrections, or subgroup restrictions. When reports from the same study provide different outcomes, the difference may be entirely legitimate rather than contradictory.
If genuinely conflicting information cannot be reconciled, document which source you use and why. Cochrane recommends establishing instructions or an agreed algorithm for handling inconsistent information across sources.
Watch Out
Designating a primary report should not erase the publication history of the study. A companion report, correction, registry record, or follow-up publication may contain information that changes how the primary article is interpreted.
06 · What This Means for You
Choose Transparently, Then Keep the Other Reports Connected
For each study with multiple reports, create a report map before making the designation. Record publication type, sample analyzed, follow-up period, outcomes, role in the study, and information relevant to your review.
A simple decision framework
If one full report clearly presents the completed study's principal analysis and relevant results
It is usually a strong candidate for designation as the primary report.
If the earliest report is preliminary or incomplete
Do not select it merely because it was published first.
If another report contains the best information for a particular eligible outcome
Use that report for the outcome while retaining the study-level primary designation.
If reports contain apparently conflicting information
Investigate the source of the discrepancy and apply a documented rule for deciding which information to use.
If no report is clearly primary
Choose the main source using explicit review-relevant criteria and document the rationale rather than pretending the hierarchy is obvious.
This process works best when the distinction between study-level and report-level information remains visible. The study record tells you what investigation you have; the linked reports tell you where particular pieces of information came from.
Once the reports are organized, you can use companion publications without double-counting participants while still benefiting from information distributed across them.
07 · A Quick Checklist
What to Check Before Designating a Primary Report
Before choosing the main report for a study, check:
That all candidate publications genuinely belong to the same underlying study.
Which publication reports the study's principal or prespecified analysis.
Which report most clearly describes the population, design, participant flow, intervention or exposure, follow-up, and principal results relevant to your review.
Whether an apparently earlier report is actually a protocol, conference abstract, interim analysis, or preliminary publication.
Whether protocols or registrations clarify which outcomes and analyses were originally planned.
Whether other reports provide unique outcomes, harms, follow-up information, methods, corrections, or other data that should be retained.
Whether apparently conflicting results can be explained by different populations, time points, outcome definitions, or analyses.
Whether the rationale for designating the primary report has been documented clearly enough for another reviewer to understand it.