Department of Psychology, Metropolitan State University, St. Paul, Minn. 55108, USA.
Psychother Psychosom. 2010;79(2):97-106. doi: 10.1159/000270917. Epub 2009 Dec 24.
Publishing separate, yet very similar pieces of a single dataset across multiple papers is known as 'salami slicing'. This practice may be motivated by researchers wishing to increase their publication counts and by the desire to increase exposure of their findings. 'Salami slicing' may also be used by the drug industry to help widely disseminate positive findings regarding its products. Journal editors across many scientific disciplines have bemoaned such duplicative publications on several occasions. However, little research has been conducted on the frequency of such publication practices, and findings have been inconsistent. No research has investigated whether 'salami slicing' may also occur in publications presenting results from pooled analyses of clinical trials.
We examined the scientific literature on duloxetine as a treatment for depression, examining how data from clinical trials were reported across 43 pooled analyses.
The vast majority of pooled analyses (88%) had at least one author who was employed by the manufacturer of duloxetine. Several pooled analyses based on highly overlapping clinical trials presented efficacy and safety data that did not answer unique research questions, and thus appeared to qualify as salami publications. Six clinical trials had their data utilized as part of 20 or more separately published pooled analyses.
Such redundant publications add little to scientific understanding and represent a poor use of peer reviewer and editorial resources.
将单个数据集的单独但非常相似的部分分别发表在多篇论文中，这种做法被称为“切片式发表”。这种做法可能是由于研究人员希望增加他们的发表数量，也可能是因为他们希望增加研究结果的曝光度。制药行业也可能会采用这种做法，以帮助广泛传播其产品的阳性研究结果。许多科学学科的期刊编辑都曾多次谴责这种重复发表的行为。然而，对于这种发表做法的频率，研究甚少，且结果不一致。尚未有研究调查“切片式发表”是否也可能发生在汇集临床试验结果的出版物中。
我们检查了度洛西汀作为治疗抑郁症的科学文献，研究了 43 项汇集分析中临床试验数据的报告方式。
绝大多数汇集分析（88%）至少有一位作者受度洛西汀制造商雇佣。有几个基于高度重叠的临床试验的汇集分析提出了疗效和安全性数据，这些数据没有回答独特的研究问题，因此似乎符合“切片式发表”的标准。六个临床试验的数据被用于 20 个以上单独发表的汇集分析中。
这种重复发表对科学理解几乎没有增加，是对同行评审和编辑资源的浪费。
