Impact Beyond the Impact Factor?
In recent years, journal impact factors have been increasingly used as a tool for quantitative analysis of the wide range of scientific and medical journals that are now available.1 Of course, there are many problems and caveats with the use of impact factors for assessing quality.2 Metrics are helpful for comparisons. However, everyone would agree that the importance of science cannot be captured by any single number or index. Nevertheless, one underlying premise behind impact factors may be somewhat useful. The more often an article is cited, the more likely that it is being read and its ideas considered and discussed, that is, the broader its potential “impact” in the field.
The impact factor of Stroke has been respectable over many years. Its most recent number is 7.041, which puts it in the upper range of most neurology and neuroscience categories. However, how does Stroke compare with other leading journals in our field? When comparing Stroke with other journals, 2 points may be important to consider. First, impact factors only capture citations within the most recent 2-year periods. Second, comparisons with other general neurology or general neuroscience journals will include many articles unrelated to the field of stroke per se. In an effort to partly normalize for these differences, we decided to look at the top 25 articles in terms of total citations (ie, most cited articles) published in Stroke from 2001 to 2010. For comparison purposes, we did the same for a few selected leading journals but only included articles that were stroke-related. By counting citations over the past 10 years, we hoped to more broadly assess the impact of these top-ranked articles. By limiting articles in other journals to only those related to stroke, we hoped to compare “apples with apples.” Reviews and guidelines were excluded. Analysis was performed in the Web of Science database. For Stroke, the following search terms were used: stroke [publication name] AND 2001 to 2010 [year published] NOT review [document type]. For comparisons, we looked at Neurology, Annals of Neurology, and Lancet Neurology, and further limited the search to stroke-related articles—stroke OR cerebral ischemia OR cerebral hemorrhage [topic]. American Heart Association guideline statements were also excluded. All resulting articles were ranked in terms of total citations, and the top 25 were pulled out (a full list of all articles is in Supplemental Table I; http://stroke.ahajournals.org). The average and SD of total citations in these top-ranked articles were then calculated. Stroke compared favorably; on average, the top 25 articles in Stroke were cited over 250 times per article over the past 10 years (Figure 1A).
A fundamental mission of Stroke is to publish all important articles in our field, including those in translational and basic science. How did Stroke fare in terms of these nonclinical science articles? We returned to the Web of Science database and tried to capture only basic science articles by including the following search terms: mice OR mouse OR rats OR rat OR neuron OR astrocyte OR oligodendrocyte OR endothelial OR glia [topic]. For comparisons, we looked at Journal of Neuroscience, Neuron, and Nature Neuroscience, and limited the search to stroke-related articles: stroke OR cerebral ischemia OR cerebral hemorrhage [topic]. Reviews were excluded, and the top 25 articles in terms of total citations between 2001 and 2010 were pulled out (full list in Supplemental Table II). Once again, Stroke fared well with total citations averaging approximately 150 per article (Figure 1B).
Of course, there are many caveats with this unofficial and surely imperfect analysis. Our search terms may not capture everything we intended to capture. The journals chosen for comparisons were selected arbitrarily. Lancet Neurology only started publishing in 2002. However, these initial findings do at least suggest that articles published in Stroke are indeed widely cited and discussed in the field. If one has an important clinical or preclinical study that deserves a wide audience, Stroke appears to be a good place to publish it.
Medical publishing and the interaction of medical journals with their readers has undergone a rapid evolution from the era of print journalism to one of combined print and electronic access of articles.3 This evolution is reflected by the decline of print subscriptions to medical journals and the rise of electronic subscriptions and electronic access of published articles. Stroke is no exception to this paradigm shift. In calendar year 2010, Stroke had 2604 combined individual and institutional print subscriptions, a decline of 351 (11.9%) from 2009. However, electronic subscriptions continued to increase and the combined number of electronic and print subscriptions for 2010 was 20 685, an increase of 386 (1.9%) from 2009. The increase in the number of total print and electronic subscriptions in 2010 reversed a downward trend seen in 2008 and likely reflected an improving global economy. Electronic access to articles published in Stroke has been increasing dramatically over the past few years. The Table demonstrates that the number of full-text article views, that is, views of an entire article not just the abstract increased by 21.4% in 2010 from the previous year to >5.6 million. Additionally, access to Stroke articles on mobile devices such as cellphones and portable electronic devices increased dramatically from 886 in 2009 to 64 622 in 2010 and will likely continue to increase exponentially over the next few years.
The implications of the favorable citation numbers for highly cited articles published in Stroke and the huge number of electronic views of the general content of the journal deserve careful consideration beyond the traditional metric of the journal impact factor. The large number of citations for leading Stroke articles suggests that these articles are considered to be important by authors who subsequently publish on topics relevant to the article previously published in Stroke. Furthermore, the frequent citation of leading articles published in Stroke supports the substantial visibility and favorable impact of Stroke on academic components of the cerebrovascular field. The astounding number of electronic accesses to Stroke provides further support for the visibility of articles published in this journal. With electronic access, a large number of readers peruse the content of Stroke on a frequent basis and presumably this written material is important to them in daily clinical practice and in the design of research projects, as well as providing help for supporting the content of their manuscripts. The standard metric of a journal's value, the impact factor, only indirectly reflects the frequency of citations and does not measure at all the hidden effect of frequent electronic accesses of articles. The authors of frequently cited Stroke articles should be congratulated for allowing us to publish their important contributions as should the prior editorial team of Stroke for providing the current team such a flourishing and highly visible journal.
The opinions in this editorial are not necessarily those of the American Heart Association.
The online-only Data Supplement is available at http://stroke.ahajournals.org/cgi/content/full/STROKEAHA.110.625541/DC1.
- Received May 11, 2011.
- Accepted May 12, 2011.
- © 2011 American Heart Association, Inc.