Editor's Choice

Setting limits to infinite error

BMJ 2008; 337 doi: https://doi.org/10.1136/bmj.a1823 (Published 25 September 2008) Cite this as: BMJ 2008;337:a1823
  1. Tony Delamothe, deputy editor, BMJ
  1. tdelamothe{at}bmj.com

    If the journal were to print an epigram over each of its sections then for Letters I’d plump for Bertolt Brecht’s description of the purpose of science. “The aim of science is not to open the door to infinite wisdom, but to set a limit to infinite error,” said the protagonist in his Life of Galileo. Setting a limit to infinite error just about sums up what Letters are for.

    All BMJ letters begin their lives as (online) rapid responses to articles, and in a generous week we select about 10% of these to publish as letters in the print journal. Our main criterion for selection is whether a response identifies an error, weakness, or incompleteness in the original article that readers would benefit from knowing about.

    This week’s Letters pages provide a good example of our aspirations. Colombians Ospina and Salcedo flag up a problem of using ibuprofen in children with viral infections, which wasn’t mentioned in the original research paper (doi:10.1136/bmj.a1767). Tassy and Retornaz from Marseilles (doi:10.1136/bmj.a1770) and Wilson from Ottawa (doi:10.1136/bmj.a1772) detect two missing elements from WHO’s report on health inequities. Laura McGlone and colleagues from Glasgow (doi:10.1136/bmj.a1774) quote work indicating that methadone in pregnancy may have adverse effects on fetuses (a paper we published in June (doi:10.1136/bmj.39514.554375.AE) was more sanguine about this).

    Sometimes the Letters pages can seem very combative, with opposing factions slugging it out. But there is a less gladiatorial route to enlightenment, which uses previous BMJ articles as “pegs” on which to hang some interesting new information. An example is the letter from Onno Terpstra, a Dutch professor of surgery, who wades into the debate over the European Working Time Directive, brandishing data (doi:10.1136/bmj.a1775). When in 1993 the Dutch parliament approved a reduction in junior doctors’ working time to 48 hours a week, “the surgical community in the Netherlands experienced all the classic stages of mourning as described by Kübler-Ross,” he wrote. A few years on, trainees are operating on the same number of cases, but working fewer hours per week. Three quarters of them are happy with the new arrangements, and applicants for training outnumber the available slots by a factor of two to three.

    A second example of interesting new information is Jim Rodgers’s discussion of how the Medical and Dental Defence Union of Scotland has been working with Terema, an organisation run by doctors and former British Airways pilots that focuses on managing the human factors in risk (doi:10.1136/bmj.a1777). And in a third example, the Critical Psychiatry Network reports that it’s possible to run an annual meeting without funding from the drug industry “albeit in less plush surroundings than usual” (doi:10.1136/bmj.a1780).

    For the past few years, we’ve been selecting responses for publication five days after print publication of the article to which they refer. That’s proved too tight a deadline for respondents with very complicated cases to assemble and for authors of criticised articles to compose their replies. For that reason we’re increasing to 12 days the time elapsing between print publication of the article and letter selection. That should allow us to notch up a few more victories over infinite error.


    Cite this as: BMJ 2008;337:a1823

    View Abstract