Department of Orthopaedic Surgery, Hospital of the University of Pennsylvania, Philadelphia, PA 19104, USA.
J Bone Joint Surg Am. 2013 Jan 2;95(1):e2. doi: 10.2106/JBJS.L.00046.
Presentations at the Annual Meetings of the American Academy of Orthopaedic Surgeons (AAOS) are often used to guide clinical practice, although many corresponding manuscripts are not published. The purpose of this study was to determine the relationship between level of evidence, rate of publication, and time to publication of presentations from an AAOS Annual Meeting.
A comprehensive literature search for all 756 studies (288 papers and 468 posters) presented at the 2001 AAOS Annual Meeting was performed to determine which of these studies were ultimately published in the peer-reviewed literature. The corresponding AAOS abstracts were each assigned a level of evidence (LOE) with use of a consensus approach and The Journal of Bone and Joint Surgery (JBJS)/AAOS LOE classification system. The rate of publication and mean time to publication for each LOE was then calculated and compared.
The overall publication rate for the 2001 AAOS paper and poster presentations was 49% after five years, and 58% after ten years. At five and ten-year intervals, respectively, 77% and 85% of Level I presentations were published; 69% and 74% of Level II presentations were published; 58% and 66% of Level III presentations were published; and 39% and 51% of Level IV presentations were published. Overall, there was a significant nonrandom difference in publication rates at both five and ten years by LOE (p values of ≤0.001). Level I and II presentations were 2.9 times more likely than Level III and IV presentations to be published after five years (95% confidence interval 1.9 to 4.5), and 2.5 times more likely to be published after ten years (95% confidence interval 1.6 to 4.0). A similar association between LOE and five and ten-year publication rates was observed for both the paper and poster subgroups. The mean time to publication was 0.9 years for Level I studies, 1.4 years for Level II studies, 2.1 years for Level III studies, and 2.7 years for Level IV studies.
The LOE of AAOS presentations is positively related to rate of publication and inversely related to time to publication. Presentations with higher levels of evidence are published in the peer-reviewed literature at a greater and faster rate than those with lower levels of evidence.
美国骨科医师学会（AAOS）年会的演讲通常被用来指导临床实践，尽管许多相关的论文并未发表。本研究旨在确定 AAOS 年会上的演讲的证据水平、发表率和发表时间之间的关系。
对 2001 年 AAOS 年会的 756 项研究（288 篇论文和 468 篇海报）进行了全面的文献检索，以确定这些研究中哪些最终发表在同行评议文献中。使用共识方法和《骨科与关节外科杂志》（JBJS/AAOS）证据水平分类系统，为每个 AAOS 摘要分配了一个证据水平（LOE）。然后计算并比较每个 LOE 的发表率和平均发表时间。
2001 年 AAOS 论文和海报演讲的总体发表率在 5 年后为 49%，10 年后为 58%。分别在 5 年和 10 年的时间间隔内，LOE I 的演讲发表率为 77%和 85%；LOE II 的演讲发表率为 69%和 74%；LOE III 的演讲发表率为 58%和 66%；LOE IV 的演讲发表率为 39%和 51%。总体而言，LOE 在 5 年和 10 年的发表率存在显著的非随机差异（p 值均≤0.001）。与 LOE III 和 IV 的演讲相比，LOE I 和 II 的演讲在 5 年后发表的可能性分别高出 2.9 倍（95%置信区间 1.9 至 4.5）和 2.5 倍（95%置信区间 1.6 至 4.0）。对于论文和海报亚组，LOE 与 5 年和 10 年发表率之间也存在类似的关联。LOE I 研究的平均发表时间为 0.9 年，LOE II 研究的平均发表时间为 1.4 年，LOE III 研究的平均发表时间为 2.1 年，LOE IV 研究的平均发表时间为 2.7 年。
AAOS 演讲的 LOE 与发表率呈正相关，与发表时间呈负相关。具有较高证据水平的演讲比具有较低证据水平的演讲在同行评议文献中的发表率更高、速度更快。
