Setting limits to infinite error

doi:10.1136/bmj.a1823

Published 25 September 2008, doi:10.1136/bmj.a1823
Cite this as: BMJ 2008;337:a1823

Editor's Choice

Setting limits to infinite error

Tony Delamothe, deputy editor, BMJ

tdelamothe{at}bmj.com

If the journal were to print an epigram over each of its sectionsthen for Letters I’d plump for Bertolt Brecht’sdescription of the purpose of science. "The aim of science isnot to open the door to infinite wisdom, but to set a limitto infinite error," said the protagonist in his Life of Galileo.Setting a limit to infinite error just about sums up what Lettersare for.

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

This week’s Letters pages provide a good example of ouraspirations. Colombians Ospina and Salcedo flag up a problemof using ibuprofen in children with viral infections, whichwasn’t mentioned in the original research paper (doi:10.1136/bmj.a1767).Tassy and Retornaz from Marseilles (doi:10.1136/bmj.a1770) andWilson from Ottawa (doi:10.1136/bmj.a1772) detect two missingelements from WHO’s report on health inequities. LauraMcGlone and colleagues from Glasgow (doi:10.1136/bmj.a1774)quote work indicating that methadone in pregnancy may have adverseeffects 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 opposingfactions slugging it out. But there is a less gladiatorial routeto enlightenment, which uses previous BMJ articles as "pegs"on which to hang some interesting new information. An exampleis 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 Dutchparliament approved a reduction in junior doctors’ workingtime to 48 hours a week, "the surgical community in the Netherlandsexperienced all the classic stages of mourning as describedby Kübler-Ross," he wrote. A few years on, trainees areoperating on the same number of cases, but working fewer hoursper week. Three quarters of them are happy with the new arrangements,and applicants for training outnumber the available slots bya factor of two to three.

A second example of interesting new information is Jim Rodgers’sdiscussion of how the Medical and Dental Defence Union of Scotlandhas been working with Terema, an organisation run by doctorsand former British Airways pilots that focuses on managing thehuman factors in risk (doi:10.1136/bmj.a1777). And in a thirdexample, the Critical Psychiatry Network reports that it’spossible to run an annual meeting without funding from the drugindustry "albeit in less plush surroundings than usual" (doi:10.1136/bmj.a1780).

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

Cite this as: BMJ 2008;337:a1823

Add to CiteULike CiteULike Add to Complore Complore Add to Connotea Connotea Add to Del.icio.us Del.icio.us Add to Digg Digg Add to Reddit Reddit Add to StumbleUpon StumbleUpon Add to Technorati Technorati What's this?