'Salami Slicing’ Found in Analyses of Antipsychotic Trials

https://goo.gl/sNwAjx

The researchers describe ‘salami slicing’ as the practice of “publishing separate, yet similar pieces of a dataset in multiple papers.” They explain that this type of duplicate publication “can distort the medical literature by making a drug appear strongly supported based on the sheer volume of publications analyzing clinical trial data.” Pharmaceutical companies often pay “key opinion leaders” to author these publications in order to lend credibility to the article and market their drugs.

Many have critiqued the practice of salami slicing when it involves multiple publications from a single study. However, less attention has been paid to salami slicing that involves pooled analyses, where researchers combine results from multiple clinical trials into one large analysis. To address this, the researchers analyzed pooled analyses for clinical trials studying the efficacy of second-generation antipsychotics for depression.

The researchers identified 43 articles, 38 of which were funded by the pharmaceutical industry. Almost 90% of the authors of these articles had financial conflicts of interest, such as being an employee of the company that manufactures the drug being studied or receiving financial support from the pharmaceutical industry. Only two articles were independent of industry funding and had no authors with conflicts of interest. Thirty-four of the articles that were industry sponsored reported using a medical writer, who are often paid by industry and used to speed up the publication process.

Results show strong evidence of salami slicing as “several such analyses did not address unique research questions and/or provided very thin slices of data.” For example, at least 9 pooled analyses were written based on the same seven studies. In addition, 21 of the articles only reported on data from one or two clinical trials, a small number given that pooled analyses are designed to combine a large amount of data from multiple trials.