No open-access location is known for this record yet.
No open-access location is known for this record yet.
I have failed over and over again in my life and that is why I succeed. (Michael Jordan, 1998) There are many more manuscripts submitted to journals than are published every year, and the number of publications is growing in a manner similar to Moore's law for computers 1. Anaesthesia receives approximately 800 submissions per year and accepts around 15% of these. We strive to review these in an independent and fair manner, selecting those which we believe are practice–changing, ethically and methodologically robust, and of interest to our readership. Manuscripts are reviewed by the Editor-in-Chief and other members of the editorial board. Approximately 20% are sent out to experts for additional external review. Manuscripts are then accepted, accepted pending revision, or declined, often with advice for the authors on how to strengthen their paper for future submission to an alternative journal. We have a rapid turnaround at Anaesthesia – authors typically receive an initial decision within 2 weeks, and articles that are accepted for publication usually appear on the journal website within 4–6 weeks of submission of the final revised paper. But how does a journal decide whether it is accepting or rejecting the right articles? The process of peer and editorial review is subjective and at risk of bias 2, and there is no central registration database, so that once a manuscript is rejected by a journal, there is no record of what happens to it. Nobody knows if it was submitted and subsequently published elsewhere, or if it was retained by the authors and never appeared in the public domain. One measure of journal success is the impact factor (IF), (http://scientific.thomson.com/isi/). This is a measure reflecting the number of times the 'average original article' in the journal has been cited in the preceding 2 years. Anaesthesia's 2015 IF is 3.8, which means that on average, original articles published in Anaesthesia in 2013 and 2014 were each cited 3.8 times in 2015. It is used as a proxy for the relative importance of a journal within its field, with journals with higher IFs deemed to be more important than those with lower ones. Impact factor is also important to career academics in terms of how they are viewed and assessed by their university employers and potential funding bodies. Higher IF journals mean a higher likelihood of their work being read and cited. Publishers also need a high IF as this affects how much they can charge institutions, and therefore how much profit they can generate. From a journal's perspective, a high IF is, therefore, important to secure reputation and ongoing funding from publishers. We decided firstly to investigate what happened to manuscripts that we rejected and what their ultimate destination was, and secondly to explore whether we were letting high-impact articles slip through our fingers. We reviewed all manuscripts submitted to Anaesthesia from January to March 2013, inclusive. We chose a historical 3 month period as we felt this gave sufficient time for all rejected manuscripts to be published elsewhere by the time of our search (December 2015). This also enabled us to assess the number of relevant citations, as this demands a 2-year window. We used Google Scholar for our search as it provides increased access to grey literature compared with Medline and retrieves more than journal articles (includes preprint archives, conference proceedings, and institutional repositories) 3, 4. We used the following search strategy: (Title) OR (Title AND Author) OR (Author AND key words). If our search was unsuccessful with the first author, we then searched with the other listed authors. We searched published literature [date of last search 21/12/2015] for the following information: has this manuscript been published (yes/no); in which journal has this been published; what was the time from submission to Anaesthesia to publication; and how many times has this paper been cited? We compared our results with papers accepted by Anaesthesia in the same time period. Anaesthesia received 172 manuscripts in the first 3 months of 2013 (Fig. 1). Of the 150 rejected papers, 87 (58.0%) were original research, 62 (41.7%) were case reports, and two were reviews. Fifty-eight (66.7%) of the original research papers were published in other journals; total citations 115, median (IQR [range]) citations 1 (0–4 [0–12]), 2.0 citations per article published. Twenty-four case reports (38.7%) were subsequently published elsewhere, with a significantly lower citation total of 31, median (IQR [range]) 0 (0–2 [0–9]), 0.8 citations per article. Twenty of the original research papers and 13 of the case reports had no citations. The delay to publication of rejected papers was median (IQR [range]) 13 (9–19 [2–30]) months. The articles were published in 40 different journals. The most common journal destination of rejected titles was the European Journal of Anaesthesiology (n = 6) however, destinations were varied and dependent on the topic and country of origin. Subspecialty journals were frequent destinations, such as Pediatric Anesthesia (n = 3), International Journal of Obstetric Anesthesia (n = 2) (Table 1). Only two articles were published as full articles in journals that had a higher impact factor in 2013 (British Journal of Anaesthesia, Journal of Thoracic and Cardiovascular Surgery); other articles were all published in journals with lower or no impact factors. During the same time-period, Anaesthesia accepted 17 original research papers, median (IQR [range]) 10 (4–14 [1–18]), 6.8 citations per article, and 5 case reports, 4 (0–7 [0–43]), 8.6 citations per article. We accepted 12.8% of original research and case reports submitted to Anaesthesia in the first 3 months of 2013. We found that 58/87 (66.7%) of rejected original research papers and 24/62 (38.7%) of rejected case reports were subsequently published in other journals 5. Half of the articles rejected by Anaesthesia were published elsewhere within 13 months, which is similar to other studies 5, 6, although the maximum time to publication was 30 months. The citation rate was considerably higher for articles published in Anaesthesia compared with articles that were rejected by Anaesthesia and published elsewhere. Most of the rejected articles were published in journals with an impact factor below that of Anaesthesia in 2013, and this is comparable with the findings of other studies 7, 8. No similar studies have been undertaken for other anaesthesia journals, although the fate of rejected manuscripts has been studied in both medical 6, 8, 9 and non-medical fields 10. Rejection from a journal does not necessarily mean that the paper is of poor quality. It may be that the scope, style or theme of the journal is different from the paper. Publication bias is a significant problem in research, and positive studies are more likely to be accepted by journals than negative studies 11. Positive studies may appear in print up to a year earlier than negative studies, and up to half of health-related studies may never be reported 12, 13. Peer review itself is a flawed process, and it has been argued that the level of agreement between reviewers is little better than tossing a coin 2. There have been various ideas on how to improve peer review. Open peer review, where the authors and readers are told who their reviewers are, has several potential advantages. It could reduce suppression of manuscripts by reviewers with a vested interest; it could encourage thorough reviews and prevent the publication of poor quality papers; and, possibly most importantly, it may convince the reviewer to be polite and constructive. However, there is no evidence that this improves the quality of the review 14; there is also the fear that open peer review would lead to a reduction in the reviewer pool. In addition, peer review is an unpaid, voluntary role that is increasingly difficult to recruit to, for the reviewing of both manuscripts and grant applications. There is also some evidence that training reviewers, and blinding reviewers to author name/institution, has not led to meaningful change 2. In light of this, alternative publishing models have been proposed. In the open access model, the author is charged a fee when a scientific paper is accepted for publication. This potentially reduces publication bias, and means that everyone who has access to the internet can access the research. However, although this works well for some journals such as PLOS ONE, this model has also been adopted by so-called 'predatory journals' 15. Here, there is a financial incentive to publish many papers quickly, and there may be limited or no peer review. Furthermore, it is much easier for author misconduct to slip through unnoticed. Whilst the medical readership has a responsibility to critically appraise evidence before acting on it, we must acknowledge the competing demands on clinician time, and that, too often, conclusions are drawn from the abstract, trusting that the science is well-conducted. Editors have a duty to alert the relevant authorities and ensure that a proper investigation is conducted if they suspect research fraud 16, and Anaesthesia has recently implemented the 'Carlisle Method' 17-20 for screening the baseline characteristics of all submitted randomised controlled trials for suspicious data not consistent with random sampling. There are flaws in using the IF as a measure of journal success – a high citation rate for a paper may mean that many people have written articles displaying their disagreement with its findings. The IF is designed to be a measure of the impact of the journal as a whole, but is frequently used to evaluate individual papers and even authors. However, the IF for a journal is often the result of one or two highly cited articles; we found that the distribution of the citations of rejected papers was highly skewed. There were no citations at all for around a third of original research and half of case reports, and the mean was pulled up by just a few highly cited articles. This mirrors the findings of Lariviere et al. who showed that 74.8% of Nature articles were cited below their impact factor of 38.1 and this was similar for Science (75.5% below its impact factor) 21. Given these limitations in IF, Anaesthesia, like many other journals, now also uses a faster and more responsive metric of the popularity of individual articles called Altmetrics (https://www.altmetric.com), which allows the impact of individual articles to be measured from the moment they are published by tracking a range of sources, including: social media (Twitter, Facebook, blogs); traditional media (news releases and reports); and use on citation/reference manager sites like Mendeley (http://www.mendeley.com). However, these measures may again only highlight the discussion around an article, and cannot be used as reliable proxies of research quality. This exercise has given us, as the editors of Anaesthesia, a clearer understanding of our quality control processes, and we would recommend this to other journal editorial boards. We believe that we accept the highest quality articles for Anaesthesia, and our articles are frequently used as the basis of further research. We hope to encourage authors of manuscripts rejected by Anaesthesia to take heart that there remains a good chance of being published in other peer-reviewed journals. The authors would like to thank Mrs Susan Jarvis-Atton from the Anaesthesia editorial office for helping with retrieving the data. AK is the Editor-in-Chief of Anaesthesia and AD was the Anaesthesia trainee Editor Fellow from 2015–16. No other competing interests declared.
Loading…
Stored only when both papers are in the corpus — 17 citations exist upstream.
Papers that cite the same references — the relational fallback while embeddings are still backfilling.
Imported from OpenAlex (CC0) — nothing here is generated.