Pleasing both authors and readers

To succeed, journals need to please both authors and readers. There is, however, a tension between the needs of the two, particularly when the authors are mostly researchers and the readers mostly practitioners. Practitioners like research articles to be short and sweet, whereas researchers want—rightly—to include enough material for critical readers (often other researchers) to be able to appraise the study and if necessary repeat it and also, increasingly, to be able to include it in a systematic review. Journals have struggled with this tension for years, and often the result is that we please nobody. Research among readers consistently shows that research articles are not well read, while many studies have shown that essential data are often missing from research reports. Now the electronic revolution offers us a chance to please both readers and authors simultaneously.

Today's BMJ includes four papers where a short version is published in the paper journal and a longer version in the electronic journal (eBMJ) (p 897-914).^1–4 We even have an acronym for the process: ELPS (electronic long, paper short). This first effort is an experiment, and we are not yet planning to introduce this system for every research article—but we may if both readers and authors are pleased.

Our experiment follows an intense debate within the journal on whether this is a good idea. The arguments in favour are those we've already advanced plus the possibility of using the pages spared for sections that are more popular than research articles. The main arguments against are that the paper journal is the “proper” journal and that not everyone has access to the internet. However, we have already said that the eBMJ is the primary journal in that it includes everything published in the paper journal and an increasing amount more.⁵ Paper and electronic versions of journals will diverge as the electronic versions exploit the full potential of the internet. The journal Pediatrics, for example, publishes some studies primarily in electronic form, with only an abstract in the paper version.

The other major concern is that some people, particularly in the developing world, do not have easy access to the internet. In the developed world access to the internet is increasing exponentially, and soon it will be accessible through television without any need for a computer. Most researchers have easy access through academic networks, and those who do not have direct access can easily obtain a copy of an article from the eBMJ in the traditional way through their medical library. In fact access to the eBMJ is probably easier than access to a print copy for anyone who is not a subscriber: since the eBMJ is free any library or other institution with an internet connection can access it immediately. Sadly, access to paper journals has been severely restricted in the developing world, and in the long run electronic forms of journals are likely to reach many more people than paper forms ever could. If we need at some stage to charge for the eBMJ then we will keep it free to those in the developing world (which we can at no extra cost to us, whereas the cost of transporting paper is substantial).

A further argument against ELPS is that it may become a licence for authors to produce interminable verbose reports. This we will resist, although reports on scientific studies may eventually expand to include sound, video, original data, software, and more. The challenge is not just to present studies in the same old way but to find ways to use the medium to full scientific advantage for both authors and readers/viewers. We accept that we could do a better job of including more essential information in paper versions of studies without necessarily making them any longer. Standard formats—such as CONSORT for publishing reports of randomised controlled trials⁶—should increase the informative value of articles, though they do generally seem to make reports longer.

As ELPS is currently experimental, we are keen to present papers in different ways. The shorter versions in this week’s journal have been made shorter by general shortening throughout all the sections of the paper, and in one case we have prepared two shorter versions, one much shorter than the other (pp 908, 912).⁴ We have, however, debated whether to increase the readability of reports by emphasising the introduction and discussion or whether to help those readers interested in critically appraising studies by concentrating on methods and results. We hope to continue the experiment by presenting further different sorts of shorter versions. We look forward to hearing views on the whole idea and on how we might best use paper and electronic media to complement each other.

Papers pp 897-914