Clinical cost of making headlines

Scientific papers should not be thrown open to the media, says Ben Goldacre, unless we want the public even more misled

Paul Broca was a French craniologist who measured brains. He was famous, and his name is given to Broca's area, the part of the brain involved in generating speech, which is often damaged in strokes. But Broca had a problem: his German brain specimens were 100g heavier than his French ones, and by rights, the French should have been superior. So Broca decided that other factors, such as overall body weight, must also be taken into account when measuring brain size: this, he said, clearly explained the larger German brains. For his prominent work showing that men had larger brains than women, Broca made no such adjustments.

Cesare Lombroso, the 19th-century pioneer of "biological criminology", also fudged the interpretation of his results: he cited insensitivity to pain among criminals and "the lower races" as a sign of their primitive nature but identified the same quality as positive evidence of courage and bravery in Europeans.

If Broca or Lombroso had failed to publish their methods and results, merely presenting their interpretations informally to the popular media, or through press releases, then the take-home messages would have been: "Science proves French people cleverer than Germans", "Science proves men cleverer than women", and "Science proves criminals are biologically inferior". These would be seductive messages to science writers even today.

So I was surprised to see Bob Ward, former head of communications at the Royal Society, speaking out in these pages against the Ingelfinger rule ("We ought to get all findings out fast", August 31). Taken at its toughest interpretation, this rule means academic journals will refuse to publish research that has been presented to the media, and with good reason.

Ward gives two explanations for why editors enforce it: first, the journals are venal and want the publicity to be attached to their own organs; second, he concedes, peer review may be an effective filter for dodgy research. But Ward called for the rule to be overturned, and for journalists to be permitted to write about unpublished data, speedily, and in the public interest.

First I have minor quibbles. Many papers are available online before print publication, rapidly after acceptance. Those with urgent clinical implications, specifically, are fast-tracked to paper publication, sometimes within a month. It is in the interest of journals to be quick, after all, to attract good contributors, and some compete very efficiently: the BMJ will reject a manuscript within a couple of weeks. And all sane editors would consider it fair to go to the media before publication if a paper contained an urgent health message.

But there is a far bigger issue at stake. It is not peer review that makes a full academic publication important: it is universal access to the full methods and results of the study. With scientific research, today as much as in the era of Broca and Lombroso, the devil is in the detail, and those details are not to be found on the pages of daily papers.

Were my 19th-century examples extreme, inflammatory and unfair? Ward's example was not so hot, either. His cited case was the "grapefruit causes breast cancer" study from July, a piece of observational epidemiology that elicited blanket media coverage.

This grapefruit paper is exactly the kind of work that is routinely misunderstood and misrepresented in the media, which seems collectively incapable of basic critical appraisal of academic work even when it is published. It was a speculative piece of observational epidemiology, and the authors themselves have been clear that it is interesting, theoretical, but highly tenuous stuff, specifically advising caution in its interpretation and implementation and flagging up the need for replication.

This is a trivial example. The archives at badscience.net are overflowing with just a small sample of the media's crimes: preposterous cherry-picking, outrageous overextrapolation, startling ignorance or whitewashing of known methodological flaws and, worst of all, reporting the authors' speculative conclusions, from the discussion section of a paper, as if they were the experimental results themselves.

The sad reality is, from the extremes of the media's MMR scaremongering to the grind of the Daily Mail 's bizarre ongoing ontological project of dividing all the world's inanimate objects into those that cause or cure cancer, the media commentariat has not earned privileged early access to scientific knowledge and information. And stories such as the media's MMR hoax (as it will come to be known) have been perpetuated by the promotion of unpublished research as if it were gold-standard work.

With the Ingelfinger rule half-relegated, popular media debate on a piece of research often takes place in a strange "post-embargo pre-publication" twilight period, where the work is available only to the media commentariat, not to doctors or academics. By the time those in the latter group get to read the paper it is too late to point out how grotesquely its significance was caricatured.

There is also a wider, gentler issue: science is a features subject, which fits uncomfortably into the "news" template at best, because it moves ahead by the gradual emergence of new themes, each supported by a raft of evidence from a range of disciplines. New discoveries - newsworthy by virtue of being unexpected - are by their nature often overturned, which is to say, they are wrong, misleading or red herrings. Often new discoveries deserve less attention, not more, and reach prominence only because of the media's obsession with unrepresentative "breakthroughs".

With medical science, bad media coverage carries a genuine clinical cost. Patients arrive at surgeries brandishing clippings, and this puts doctors in an extremely difficult position. To take the clipping, nod sternly over half moon specs and explain that newspapers are full of "lies" is inappropriate, dismissive and injurious to the doctor-patient relationship.

But to practise medicine based on the content of newspapers is an equally absurd suggestion. Doctors need access to original research papers so they can critically appraise their methods, results and relevance to clinical practice. When the media's coverage - especially the more mendacious and misleading material - is based on informal personal briefings, half-overheard conference presentations, industry press releases or pure fantasy, this is impossible.

If it is accurate coverage we are after, then the media should cover only research that has been published in full - in any format where the methods and results are fully available, regardless of "peer review" - and with a prominent link to the primary source. If journalists are sure a piece of research is relevant enough to fit into the template of an eyecatching headline, then they should be happy to see their artistic interpretations displayed alongside the original.

And of course it would be great if all journals moved quicker. But the media has not earned any privileged access to data; if anything it has abused it.

Ben Goldacre is a writer, broadcaster and medical doctor who works full-time for the National Health Service.