Percentage of Studies Presented at a Conference Prior to Publication in Emergency Medicine Journals.

Abstract

Traditionally, authors will present a manuscript at a conference prior to submitting the subsequent manuscript for consideration in a journal. While peer-reviewed publications are typically considered to be the criterion standard in research,1 presentations of abstracts at conferences have several distinct advantages. First, presentation of abstracts at a conference offers an opportunity for early feedback and review, which can help identify issues prior to submission. This can allow authors the ability to refine their study question or revise their study design prior to completion of the project. In addition to this prepublication feedback, conference presentation can also provide an opportunity to identify future collaborators and gain early presentation experience for more novice researchers. Moreover, the average time delay between conference presentation and eventual publication is 1 to 3 years,2, 3 so this may allow for an early route for dissemination. However, this must be balanced with increasing time and funding limitations as well as alternate dissemination avenues through social media. In light of these factors, it is important to determine the degree to which manuscripts are presented prior to publication and which factors influence this decision. The first step in this process is to quantify the degree to which this occurs. To the best of our knowledge, there are no data demonstrating what percentage of manuscripts are actually presented as abstracts prior to being published. This was a cross-sectional, observational study that sought to determine the percentage of manuscripts that were presented at a conference prior to publication and the distribution by publication year, conference type, and study design. The study was conducted in accordance with best practice recommendations for chart reviews.4 Two investigators independently reviewed all publications from January 2009 to December 2018 from two major Emergency Medicine (EM) journals (Annals of Emergency Medicine and Academic Emergency Medicine). These journals were selected because they had the highest impact factor among EM journals and explicitly require authors to disclose any presentations of the work prior to publication. All original research articles were included. Narrative reviews, case reports/images, letters to the editor, and consensus conference proceedings were excluded. A data extraction tool was developed, piloted, and modified in accordance with the pilot data. Two investigators independently dual-extracted the data into the data collection form for all included studies. When less than five manuscripts were published from a given study design, they were placed into the “other” category. Any discrepancies were resolved by consensus with the addition of a third investigator as needed. Data were grouped by year of publication, study design, and distribution of conference types. Data are presented primarily as descriptive statistics including means and standard deviations (SD). Statistical analyses were performed using Microsoft Excel (version 16.1). A total of 2,409 articles met our inclusion criteria (comprising 1,414 from Academic Emergency Medicine and 995 from Annals of Emergency Medicine) with 1,192 (49.5%) having been presented as conference abstracts prior to publication. Of those that were presented, the mean (±SD) number of presentations was 1.4 (±0.7). The rate of conference presentations decreased over time, with 57.0% of manuscripts presented in 2009 versus 38.8% in 2018 (Figure 1). However, the mean number of presentations per manuscript was relatively stable over the 10-year period, ranging from 1.2 to 1.5 (Data Supplement S1, Figure S1, available as supporting information in the online version of this paper, which is available at http://onlinelibrary.wiley.com/doi/10.1111/acem.13911/full). Of the total conference presentations, 89.6% were presented at a national or international conference, 8.3% were presented at a regional conference, and 2.0% were presented at a local conference. Overall, 61.7% of presentations were at an EM conference and 38.3% were at non-EM conferences. Among non-EM conferences, the most common specialties were pediatrics (33.2%), toxicology (5.6%), and cardiology (5.2%; Data Supplement S1, Figure S2). When assessed by study design, the most commonly presented studies were before–after studies (70.8%), prospective observational studies (55.3%), and randomized controlled trials (51.6%), while the least common were consensus/Delphi analyses (31.6%) and systematic reviews (27.6%; Data Supplement S1, Figure S3). This study identified several interesting findings that warrant further discussion. Overall, only half of all manuscripts were presented prior to publication with a notable decline over time. This may be influenced by the increasingly prominent role that peer-reviewed publications have played within the promotion and tenure process.1 Because the time from abstract presentation to manuscript presentation can be 1 to 3 years,2, 3 it is possible that authors may opt to forego abstract presentation in order to shorten the time to publication. Additionally, there can be high costs associated with attending a national or an international conference. A recent survey found that the average funding for physician continuing medical education ranges from $3,500 to $5,000 per year.5 This may limit the ability of physicians to attend more than one or two conferences each year. Moreover, competing time commitments may further restrict their ability to present to specific times of the year. Furthermore, with the rise of social media, there are new avenues of collaboration and dissemination pathways that were previously limited primarily to conference attendance. This may also explain the decreasing number of conference presentations over time, which is inversely correlated with the rise of social media. However, it is important to weigh this against the potential for unintentional consequences, including decreased networking opportunities and reduced ability for junior researchers (both residents and faculty) to gain early presentation exposure.6 Another interesting finding was that over one-third of abstracts published in EM journals were presented at non-EM conferences. We found that pediatric conferences represented the most common non-EM conference type occurring nearly six times more than the second most frequent conference. This may be reflective of the increasing number of publications seen within pediatric emergency medicine over the past 10 years.7 Other fields, such as toxicology, cardiology, and neurology were also represented in 3% to 5% of cases. This may be due to increasing expansion of EM research into areas typically dominated by other fields. Importantly, presenting at non-EM conferences provides an opportunity for sharing research outside of EM with the associated fields, as well as opportunities to expand researchers’ networks by attending these conferences. Finally, we found that before–after, prospective observational, and randomized controlled trials were the most frequently presented, while retrospective, consensus/Delphi, and systematic reviews were less frequently presented. This is an interesting finding, as those most likely to receive modifiable feedback were least likely to have the opportunity to receive presubmission feedback. With prospectively collected data, it may be more challenging to incorporate feedback after a study has already began. However, because authors often present preliminary data prior to study completion, presenters may identify major design flaws and redesign (and potentially restart) their study to better address these issues before the full study is completed. This study also has several limitations to consider. First, it was performed in two EM journals and may not reflect the full spectrum of EM research. It is possible that the true incidence of prepublication presentations may be higher or lower than the present study if assessing all EM journals. However, since most journals do not explicitly require that authors list any prior presentations, this was not feasible to perform with most other journals. It is also possible that some authors may have presented their work without disclosing it to the journals. Additionally, the present study is limited to only EM literature and may not reflect that of other fields. Future studies should determine whether this finding is consistent in other fields. Moreover, this study was retrospective in nature and subject to all the inherent limitations of this approach. Finally, the present study does not explain why many authors did not present their research prior to publication and future studies should evaluate this component. In conclusion, less than half of all original research manuscripts were presented at a conference prior to publication, with a progressive decline in the number of manuscripts presented over time. Future efforts should identify the underlying reasoning for not presenting at a conference and why this trend has been increasing over time. Data Supplement S1. Supplemental material. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.