‘Salami slicing’ yields many articles from one trial

As discussed by Toronto Star‘s Stuart Laidlaw and Pharmalot’s Ed Silverman, the journal Psychotherapy and Psychosomatics recently drew attention to the practice of publishing multiple journal articles from the results of one clinical trial in a study they titled “A Case Study of Salami Slicing: Pooled Analyses of Duloxetine for Depression” (PDF). Duloxetine, for the record, is an Eli Lilly drug better known as Cymbalta.

Photo by huntz via Flickr.

Salami slicing is well known (and disliked) by top-flight journal editors who, Laidlaw said, often require authors to “declare if any part of the underlying research has been – or soon will be – published elsewhere.” Still, the researchers behind the Psychotherapy and Psychosomatics paper found it remains widespread, at least in the case of Cymbalta.

The authors speculate that the practice is propelled by two key factors: researchers who want to increase their exposure and publication counts and pharmaceutical manufacturers who want to “widely disseminate” the results of positive studies and create an artificially inflated base of scholarly support for their products.

Researchers defend the practice as cost-effective, saying that medical trials can be prohibitively expensive, and salami slicing is one way to extract as much research as possible from one trial. Additionally, they say that trials are intentionally designed to be sliced, and that there are often multiple separate issues that each warrant their own manuscript.

Laidlaw also found at least one researcher who took issue with the semantics of the practice’s sensation-friendly label.

“I would be careful of anything that’s given a folksy name,” says Dr. Ralph Meyer, director of the National Cancer Institute of Canada’s Clinical Trials Group at Queen’s University.

Leave a Reply