Yesterday morning the BMJ’s press officer needed to locate a rapid response about Tamiflu from Peter Doshi, a postdoctoral fellow at Johns Hopkins University, Baltimore. Doshi’s response accuses the drug’s manufacturer Roche of “all talk and no action” following its promise to share full clinical study reports (CSRs) for 10 treatment trials.
Our Tamiflu open data campaign page was a logical starting point. It wasn’t there. Doshi had in fact responded to this news story, posted last week: Roche says it will not relinquish control over access to clinical trial data.
Our press officer’s problem was doubtless one experienced by many bmj.com readers. How do you find a response’s online location? Arguably we make it harder for readers by often commissioning three or four articles on a similar topic. Head to head articles are a classic example. Our last two head to head articles asked, “Should the NHS work at weekends as it does in the week?” Yes, argued NHS medical director Bruce Keogh. Paul Flynn, chair of the BMA’s consultants committee, disagreed.
Keogh’s article received 42 responses. Flynn’s received three. Surely it would be better to have a single response stream, sparing readers the inconvenience of accessing both articles to read what people are saying?
We can turn off commenting on specific articles, but don’t tend to, unlike other publications, such as The Guardian, according to Tom Happold, its former head of multimedia, whom I met last week. Should we disable responses on some articles, so the debate is concentrated in one place?
The head to head issue could be easily solved by posting these as a single article. We have considered this in the past. But it might be tricker for more complex bundles of articles. The BMJ investigation into sports drinks, for example, published last July, included six features and an editor’s choice. The main feature attracted 19 responses, another attracted five, and the remaining four one each. The editor’s choice also attracted one.
The wisdom of the crowd suggests that readers are logically drawn to a main article. Supplementary articles seem to raise legitimate questions, and so it would be wrong to deny readers the opportunity to respond. And responses are searchable. But does this serve our readers, many of whom might prefer to read a single thread of comments rather than having to switch between articles?
David Payne is editor, bmj.com, and readers’ editor