Department of Neurosurgery, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts, USA.
PLoS One. 2012;7(8):e44327. doi: 10.1371/journal.pone.0044327. Epub 2012 Aug 31.
Medical studies are more likely to report favorable findings when a conflict of interest is declared. We aim to quantify and determine the effect of author disclosure of conflict of interest on scientific reporting.
Abstracts from an international spine research meeting (North American Spine Society 2010) were selected that specifically evaluated a device, biologic, or proprietary procedure. They were then made anonymous to reviewers. An item of interest was established in each of the abstracts in order to standardize evaluation. Next, three blinded reviewers independently rated the abstracts as favorable, neutral, or unfavorable with regard to the item of interest. Additionally, the blinded reviewers attempted to predict whether a related disclosure was made. The meeting disclosure index was used to tabulate the minimum US dollar value attributable to disclosures.
Of the 344 total abstracts, 76 met inclusion criteria. In 79%, a related conflict of interest was reported. The amount of the disclosure was incompletely reported in 30% of cases. Where available, it averaged a cumulative minimum of $219,634 USD per abstract. The results of the abstracts were judged to be favorable, neutral, and unfavorable in 63%, 32% and 5% of abstracts, respectively. There was no correlation between the presence of a related disclosure and the findings of the studies (p = 0.81), although interpretation of this is limited by a small sample size and an overall apparent bias to report favorable studies. Additionally, the blinded reviewers were unable to predict whether a related disclosure was made (p = 0.40).
No association existed between the presence of a related disclosure and the results of the studies. While the actual compliance with reporting a potential conflict of interest is unable to be determined, the value amount related to the disclosures made was inadequately reported according to meeting guidelines.
当披露利益冲突时，医学研究更有可能报告有利的发现。我们旨在量化和确定作者披露利益冲突对科学报告的影响。
选择了国际脊柱研究会议（北美脊柱协会 2010 年）的摘要，这些摘要专门评估了一种设备、生物制剂或专有的程序。然后，这些摘要被匿名给评审员。为了标准化评估，在每个摘要中建立了一个感兴趣的项目。接下来，三位盲审评审员独立地根据感兴趣的项目对摘要进行了有利、中性或不利的评价。此外，盲审评审员试图预测是否做出了相关披露。会议披露指数用于计算可归因于披露的最低美元价值。
在 344 篇总摘要中，有 76 篇符合纳入标准。在 79%的情况下，报告了相关的利益冲突。在 30%的情况下，披露的金额未完全报告。在可用的情况下，每份摘要的平均最低披露金额为 219634 美元。摘要的结果被判断为有利、中性和不利的分别占 63%、32%和 5%。存在相关披露与研究结果之间没有相关性（p=0.81），尽管由于样本量小且总体上存在偏向报告有利研究的情况，因此对这一结果的解释受到限制。此外，盲审评审员无法预测是否做出了相关披露（p=0.40）。
存在相关披露与研究结果之间没有关联。虽然无法确定实际报告潜在利益冲突的遵守情况，但根据会议指南，与披露相关的金额报告不足。
