Duplicate Publication and Salami Slicing: When One Study Becomes Many Papers

Evidence synthesis assumes each study counts once. Duplicate publication reports the same data in more than one place, which can double-count it and inflate its weight in a meta-analysis. Salami slicing splits one study into the thinnest publishable fragments. Both are legitimate only in narrow, disclosed circumstances, and both can quietly bend the summary of a field.

Why repeating data corrupts the evidence base

Evidence synthesis assumes that each study counts once. Meta-analyses pool results as though every included trial is a separate piece of information. When the same data appear in two or three papers under different titles, that assumption breaks, and the data get counted more than once, giving one study inflated weight in the pooled result.

So duplicate publication is not merely untidy or a matter of etiquette. It can quietly bend the summary of an entire field by letting a single set of patients vote several times.

Duplicate and redundant publication

Duplicate publication means reporting the same study, or a substantial part of it, in more than one place without disclosure. Redundant publication is the broader term for work that overlaps substantially with something the authors already published. The editorial recommendations single this out because double-counting data or over-weighting one study's results is a real risk to readers and to synthesis.

There is also a simple transparency rule underneath it. Authors are expected to tell editors about related work and prior publications, and to provide copies so the editor can judge the overlap. The problem is rarely the overlap alone; it is overlap that is hidden.

Salami slicing: one study, many papers

Salami slicing is the practice of splitting a single body of research into the thinnest publishable pieces, each reporting a fragment. One well-designed study that could be one strong paper becomes several thin ones, timed and framed to look independent.

The incentive is easy to see in a culture that counts publications. The cost lands on readers, who now have to reassemble a scattered study, and on synthesis, which may mistake the slices for separate studies. There are legitimate reasons to publish a large project in parts, but the test is whether each part stands on its own and whether the relationship between them is disclosed.

Text recycling and the gray zone

Not all reuse is equal. Text recycling, sometimes called self-plagiarism, is reusing one's own previously published wording in a new paper. Some of this is unavoidable and harmless; a methods section describing an identical procedure can only be phrased so many ways. Editorial guidance treats it as a spectrum rather than a simple offense, distinguishing minor reuse of methods language from recycling substantive results or discussion.

The honest position is that this is a gray zone with reasonable boundaries, not a bright line. What keeps it on the right side is disclosure and proportion: reusing necessary technical description is different from passing off old analysis or conclusions as new.

When a second publication is legitimate

Publishing the same work twice is not always wrong. Secondary publication can be entirely proper when specific conditions are met: both editors agree, an interval respects the original's priority, the second version targets a genuinely different audience, it faithfully reflects the original data and interpretation, and it clearly tells readers it is a republication of work reported elsewhere.

Translation into another language for a different readership is the classic example. The difference between this and duplicate publication is not the repetition itself. It is consent, transparency, and honest labeling.

How a reader and a synthesizer get misled

For an ordinary reader, the risk is modest but real: mistaking three papers for three independent confirmations when they describe one study. A tell is heavy overlap in the authors, the sample size, and the setting across papers that never cite one another as the same underlying dataset.

For anyone doing a systematic review, the stakes are higher, which is why careful reviewers actively hunt for duplicate data, checking sample sizes, recruitment dates, and trial registration numbers to make sure one cohort is not entering the analysis twice. The safeguard against sliced and duplicated data is the same in both cases: trace the papers back to the studies they actually came from.