The congress was a great success: Yes, but what about research?

There are 3353 orthopedic surgeons and 980 orthopedic surgery residents in Turkey (Halit Pınar, MD, President of the Turkish Orthopedics and Traumatology Society [TOTBİD], oral communication, December 2019) and 807 are board-certified (TOTBİD Head Office, oral communication, November 2019). TOTBİD, with 3220 members is one of the largest orthopedic associations in the world and the National Turkish Orthopedics and Traumatology Congresses are being held by this association since 1966. This congress offers opportunities to orthopedic surgeons, physiatrists, rheumatologists, physiotherapists, basic scientists, and many others to present, share, and discuss their research in musculoskeletal medicine and gain useful feedback and comments from the scientific community. This meeting also serves to support and encourage the development of new research projects by young researchers.

The 29^th congress was recently held between 22–27 October, 2019 in Antalya with 1750 attendants (TOTBİD Head Office, oral communication, November 2019). Apart from plenary lectures, round-table discussions, conferences, industry exhibitions, friendly gatherings, and prolonged dinners outside under the lovely Antalya night-sky, there were 307 oral and 277 poster presentations totaling to 584 in the 29^th congress (1). Everybody was happy, there were emotional farewells during departure, and the congress was a total success.

Or was it not? Are high numbers of attendance and substantially high numbers of presentations enough to call a congress successful?

Previously, two papers were published in this journal in 2013 and 2019 (2, 3), unsurprisingly leading to no debate. The first one set off to evaluate the fate of abstracts presented at the 20^th Turkish National Orthopaedics and Traumatology Congress in 2007; first to determine the publication rates of full-text articles in peer-reviewed and PubMed indexed journals after presentation of abstracts at the congress, secondly to determine the time lag from the congress date to publication of full-text articles, and thirdly to assess the consistency between abstracts and the subsequent publications (2).

Figures were not very satisfactory; publication rates from the 20^th congress showed that only 29.5% (227/770) of abstracts presented were ultimately published in peer-reviewed and PubMed indexed journals over the subsequent 5 years. Besides, 14.5% of the 227 published articles were published prior to the presentation at the congress and only 22% of the 506 poster presentations were ultimately published. Eighty per cent of the articles showed changes from their presentation abstracts, 48% of the published articles had major inconsistencies like discrepancies in the study objective/hypothesis, sample size, and primary outcome measure.

The objective of the second paper (3) was to determine the publication rates of abstracts presented at the 23^rd (2013) and the 24^th (2014) National Turkish Orthopedics and Traumatology Congresses in PubMed indexed journals and compare this data with the publication rates published previously from the 20^th congress (2007) (2). Results from this study was not very promising either. Of the 993 presentation abstracts from the 23^rd congress and of the 940 presentation abstracts from the 24^th congress, 28% and 24.9% were followed by a full-text article in peer-reviewed journals indexed by PubMed, respectively. The rates of publication of the poster presentations were 23% from the 23^rd and 18.6% from the 24^th congresses. Again, 18% abstracts from the 23rd congress and 15.8% abstracts from the 24^th congress were published as full-text articles prior to the presentation at the congress.

Surprisingly, publication rates from the 23^rd and the 24^th congresses were even lower when compared to the 20^th congress. In short, we had not improved, hence the title of the article.

A time-span of 4 years for both the 23^rd and the 24^th congresses might be regarded as insufficient for full-text articles to get published. However, in similar studies in the field of orthopedics, the mean time to publication were 17.6 and 15.6 months (4, 5). One study found that although the total number of published presentations increased each year, the likelihood of publication decreased after the third year (6). The previous study analyzing the 20^th congress also found that the mean time to publication was 14.9 months and 73.9% of the presentations were published within the first 2 and 88.5% in the first 4 years (2). Consequently, a time span of 4 years is sufficient to have a quite conclusive idea about the publication rates of meeting presentations and a longer time span will not significantly change the publication rates.

Why is it so important to get a meeting presentation published then? First of all, publication in a peer-reviewed journal is the gold standard for presenting research to a wide scientific audience. Secondly, it is simply the best indicator of the real scientific success of a meeting, because the quality of an abstract is associated with the likelihood of subsequent publication that pass the more extensive review process of a scientific journal (2–4, 7). Therefore, it is important to question the scientific validity of the scientific meetings through the assessment of the abstract-to-publication rates in the peer-reviewed literature. The publication rate potentially serves as a quality benchmark for the meeting (8).

Reports across medical specialties have documented that a considerable number of presentations never reach the ultimate goal of publication, with rates ranging from 22 to 89% (2, 4). However, the 28% and 24.9% subsequent publication rates from the 23^rd and the 24^th congresses were quite disappointing (3).

Although it is known that podium presentations typically consist of studies of higher scientific value and they are more than twice as likely to be published compared to poster presentations, 23% and 18.6% publication rates for poster presentations still suggest that the scientific quality of the poster presentations at the 23^rd and 24^th congresses were irrefutably insufficient (2, 3).

One reason for this low publication rate may be partly due to the high number of abstracts accepted for presentations. In 2007, out of the 791 abstracts submitted 773 (97.7%) were selected for presentation while only 18 (2.3%) were rejected (2). In 2009, out of the 864 abstracts submitted 793 (91.8%) were selected for presentation, and only 71 (8.2%) rejected (2). In 2011, out of the 968 abstracts submitted 866 (89.5%) were selected and only 102 (10.5%) were rejected (2). In 2013, 993 out of the 1242 (80%) submitted abstracts and in 2014, 940 out of the 1135 (82.8%) were accepted for presentation (3). In 2019, out of the 647 abstracts submitted 584 (90.3%) were selected and only 63 (9.7%) were rejected (TOTBİD Head Office, oral communication, November 2019).

Ideally, the abstracts selected for presentation should be of high quality. Although the number of abstracts submitted for the annual meeting for the American Academy of Orthopaedic Surgeons (AAOS) grows continuously, AAOS keeps overall acceptance around 25% each year, equally split between oral and poster presentations (6). Although a tendency towards increasing rejection rates is observed, 80 to 97.7% acceptance rates in the Turkish congresses are still in great contrast with the AAOS rates, aside from the fact that most of the scientific work in the world is done in United States of America.

Maybe, a more realistic approach would be comparing the Turkish congresses to the European Federation of National Associations of Orthopaedics and Traumatology (EFORT) meetings. EFORT is an organization in which Turkey actively takes part and Europe is the socioeconomic community and geographic region that Turkey locates herself to. The publication rate for the EFORT congress in 2011 was 41.7% (7), whereas the publication rates for the 2013 (3) and the 2014 (3) congresses were 28% and 24.9%, respectively. Still, with the 41.7% publication rate, the 2011 EFORT congress is in the middle of the range reported in the literature (7). Furthermore, in a recent systematic review, the publication rate of 307,028 abstracts presented in biomedical meetings was 37.3%, which was interpreted as low (9).

Another reason for the low publication rate is probably due to the considerably high number of case report abstracts among poster presentations. It is a fact that scientific journals indexed by SCI or SCI-E publish very few case reports and some of them do not publish case reports at all and currently getting a case report published in a respectable journal is quite troublesome (10). Therefore, abstract-to-publication rate was very low with case reports affecting the overall publication rates. It is known that oral presentations, experimental studies, studies with higher levels of evidence, randomized studies, prospective studies, and multicenter studies are more likely to achieve subsequent publication (7). On the other hand, we share the concerns of the Congress Organizing Committee that reducing the number of accepted presentations may lead to a decrease in the number of participants.

On the other hand, we believe that most of the abstracts were never transformed into full-text articles. It is known that a lack of time for manuscript preparation (11) and publication not being an aim or being a low priority (8) are main barriers to publication. We find this reluctance for publishing despite the well-known pressure on physicians with academic ambitions is surprising. A probable reason for this reluctance may be due to the fact that it is still possible to get a maximum of 10 points out of the 100 needed for associate professorship application with oral presentations even if they are not published as full text articles (12).

Consequently, in order to reach a satisfactory level of abstract- to-publication rate, abstracts submitted to National Turkish Orthopedics and Traumatology Congresses should undergo more rigorous peer-review and more demanding guidelines for acceptance should be adopted. Implementing a more standard and structured abstract format that requires authors to provide the necessary study details, designating the level of evidence of the research and indicating the type of research (randomized trial, observational study, review, case report, survey, or basic science) described by the presentation abstract may improve overall quality (2, 3). Of course, the abstract-to-publication rate is not the sole evaluation criterion for the success of a congress, but it indeed is the single measurable variable for the research activity presented at a meeting.