BMJ papers audit

Prepared by Sara Schroter, senior researcher, The BMJ

The BMJ receives over 3000 research paper submissions each year (Table 1) but only accepts a very small proportion. It is The BMJ’s vision to be the world's most influential and widely read medical journal, and we also aim to publish material that will help doctors to make better decisions. It is difficult to measure the influence of a medical journal, but we feel that there has been too much emphasis on the number of citations that papers generate and journal impact factors. We want our research papers to be scientifically valid, highly visible, widely read, clinically relevant, and of interest to international readers, as well as being frequently cited. As such, we now track a series of indicators for our published research papers. By collecting and reporting data routinely we can:

1. Evaluate the effects of changes in our editorial policies e.g. we can assess whether we are attracting papers that are well read and clinically relevant.
2. Become better informed about the type of research we publish e.g. are we publishing more international research and is it reflective of the areas we want to attract, etc?
3. Show prospective authors average citation and download figures to encourage submissions.
4. Identify our "best" papers - what are the characteristics of papers that are most widely read, cited, and picked up in secondary sources, etc.
5. Routinely assess our turnaround times for published research papers enabling us to identify any bottlenecks, etc.

Table 2 presents data for all The BMJ's research papers published between 2004 and 2012 (excluding the Christmas issues). It includes the following indicators:

• Total number of accesses on bmj.com (abstracts + full text + PDF) in first 3 months of publication
  Pickups in secondary sources

* Evidence Updates

* Journal Watch (General Medicine)

* Evidence Based Medicine journal

* Evidence Based Nursing journal

* Evidence Based Mental Health journal

* Faculty 1000 Medicine

• Rapid responses in first 12 months after publication (we acknowledge that not all rapid responses are positive but regard this measure as an indicator of visibility. We have excluded authors’ responses from the counts).
• Number of aggregate citations in year of publication plus one, two, three, four, five calendar years

Table 2 also includes the proportion of international research and various study designs, and average turnaround times:

• Proportion of international research (based on address of corresponding author)
• Proportion of papers reporting randomised controlled trials
• Proportion of papers reporting systematic reviews and meta analysis
• Interval in days between acceptance to online publication