Website hits and particularly pdf downloads provide direct evidence of readers' interest in papers published in journals. Tobacco Control's website has allowed examination of web hits and downloads each month and cumulatively since March 1998 (issue 7‐1).
In March 2006, we examined website data and citations shown on the Institute for Scientific Information's Web of Science for all 553 original articles, reviews, editorials and special communications published in Tobacco Control and its peer‐reviewed supplements from issue 7‐1 till issue 13‐2, comparing press‐released and non‐released articles. Articles published subsequent to 13‐2 (June 2004) were not examined because publication lag times would have meant there would have been few citations to papers published after that time.
Press releases were issued to over 1000 media outlets around the world by the BMJ's press office for 47 original articles published during the study period (table 1).
Table 1 Press releases issued by the BMJ's press office to over 1000 media outlets around the world.
| Mean website hits | Mean pdf downloads | Mean citations | |
|---|---|---|---|
| Press released (n = 47) | 7430 | 969 | 13.98 |
| Non‐press released (n = 506) | 3227 | 393 | 6.676 |
| Total (n = 553) | 3584 | 442 | 7.297 |
Press‐released papers received 2.3 times more web hits than non‐press‐released papers (p<0.001), 2.5 times as many pdf downloads (p<0.001), and were 2.1 times more likely to be cited (p<0.001). Eleven papers (23.4% of those press released) which received more than 20 citations (range 21–90) in the sample period accounted for 58.6% of all citations for press‐released papers.
Papers are selected for press release because of their anticipated newsworthiness. Newsworthiness is a subjective quality that reflects staff and editor's judgements about the likely interest that journalists will have in a paper's findings. It is not a judgement that is necessarily governed by the “importance” of a paper to the research community. When this judgement is accurate and a press release stimulates widespread news coverage, literally hundreds of millions of people globally may be exposed to the story, some of whom will have personal or professional interests in wanting to then locate and read the research article. A recent paper by SC on the effect of Kylie Minogue's breast cancer diagnosis on mammography screening1 received coverage in over 950 news outlets, including the Chinese People's Daily and Pravda.
This study design does not allow anything more than speculation about whether it is a paper's contents or the fact it has been press released which is responsible for the more than doubling of web visits, downloads and subsequent citations. However, Philips et al showed that research articles published in the New England Journal of Medicine which were reported in the New York Times received 72.8% higher citations in their first year after publication than articles not reported in the newspaper. Their study included a 3‐month period during which the New York Times was on strike but still produced an undistributed “edition of record”. Articles covered by the newspaper in that 3‐month period were no more likely to be cited than articles not reported by the newspaper.2
Footnotes
Competing interests: None.
References
- 1.Chapman S, McLeod K, Wakefield M.et al Impact of news of celebrity illness on breast cancer screening: Kylie Minogue's breast cancer diagnosis. Med J Aust 2005183247–250. [DOI] [PubMed] [Google Scholar]
- 2.Phillips D P, Kanter E J, Bednarczyk B.et al Importance of the lay press in the transmission of medical knowledge to the scientific community. N Engl J Med 19913251180–1183. [DOI] [PubMed] [Google Scholar]