“Jeez, this is a bit like having a serious conversation on top of a mountain fifteen yards apart in 90mph winds,” tweeted doctor and journalist Ben Goldacre (@bengoldacre). “Yes, and I’m a lousy typist to boot!” hollered back Randy Schekman, Nobel laureate and editor of open access journal eLife (@elife). They were attempting a debate on Twitter (#benandrandy) about scientific publishing, following Professor Shekman’s provocative assertion that “science must break the tyranny of the luxury journals.” It did work, sort of, and other tweeters joined in. E-life summed it up later using Storify, an app that gives you a blank sheet and a feed of tweets, facebook posts, and other links to curate into a linear account; a story. A week or so later Richard Smith (@Richard56) and I (@trished) gave it a go, in a live Twitter debate about our recent Head to Head article: Should journals stop publishing research funded by the drug industry? Like all BMJ Head to Heads, this had been commissioned as two independent halves, with Richard Smith and Peter C Gøtzsche saying yes and me saying no, but no direct to-and-fro between us. BMJ readers had pitched in, with some great Rapid Responses, 3000 downloads of the linked podcast, and a close run poll—786 readers (55%) voted for medical journals to stop publishing research funded by the drug industry. The article got tweeted widely on publication, and it continues to reach thousands of readers as Altmetrics and bmj.com download data show. So would a live Twitter debate between the authors help to tease out the arguments, and would we manage to organise the chaos? Here’s how we ran it:

• The published article and BMJ podcast included the footnote: “Join the authors live on Twitter to debate the issue on 21 January, 1200-1230 GMT at #pharmaban.”
• Richard Smith and I both tweeted the article’s link and #pharmaban a couple of times in the days leading up to the Twitter debate.
• an hour or so before the start of the Twitter debate Richard and I both drafted and saved tweets each giving the three main points of our opposing views.
• I drafted a few tweets to send just before the debate—linking to the article; linking to the Rapid Responses on bmj.com; and explaining the rules of engagement—and I tweeted these just before noon.
• at noon Richard sent his three preloaded tweets giving main point, then I tweeted mine. Then the free debate opened.
• at 12.30 Richard and I stopped, although a few people continued to debate without us.
• by serendipity, our Twitter debate coincided with the lunch break at the annual European conference of the International Society for Medical Publication Professionals ISMPP. We learned about this via a tweet a few days before the debate, and sent a pile of pdfs of the article to the BMJ/BMJ Open booth at the conference. Many ISMPP delegates joined in.
• Here’s our Storify summary of the debate.

Did it work? And should the BMJ make Twitter debates a regular feature, at least for Head to Heads? Trish Groves, head of research, The BMJ. Twitter @trished