For healthcare professionals only

Editor's Choice

A commentary on commentaries

BMJ 2004; 328 doi: (Published 22 April 2004) Cite this as: BMJ 2004;328:0-f
  1. Richard Smith (rsmith{at}

    Doctors don't worry whether they will continue to have jobs. No matter how successful they may be, illness, disease, pain, and suffering will continue. Medical publishers, in contrast, fret about the future. The people who make their living in the space between authors and readers—that is, editors, publishers, librarians, and various other more mysterious types—may all be rendered redundant as authors go directly to readers courtesy of the world wide web. One defence we have is peer review, but it's a poor, shivering beast—albeit a holy one. Another defence is to “add value,” and this is one reason why we publish commentaries—wise words that we commission to complement the original studies that are submitted to us for free. Perhaps because our anxiety about our future is increasing we have in this issue three commentaries, one of which is almost as long as the study on which it comments.

    Commentaries often pick up on particular issues—perhaps statistical, ethical, or legal—raised by studies. Occasionally they offer different perspectives from those of the authors, coming perhaps from patients. Sometimes they set the studies in context, although this is more the job of an editorial. Rarely they may severely criticise studies, and we will have agreed to publish the study only if it is accompanied by a commentary.

    Terry F Pechacek and Stephen Babb comment (p 980) on the study that shows that a public smoking ban reduced admissions for myocardial infarction in a small, isolated town in Montana (p 977). They point out weaknesses in the study: the absence of data on actual exposures to secondhand smoke; the small size; and the unexpectedly large effect. But the main point of their commentary is to draw readers' attention to the increasing evidence that small exposures to tobacco can cause large increases in the risk of cardiovascular disease.

    The commentary by Paul Ruddock addresses that age old medical question of what is normal (p 987). A study from Korea suggests that patients with “high normal” serum aminotransferase concentrations may be at increased risk of liver disease compared with those with “low normal” concentrations (p 983). Korea has high levels of liver disease, and Ruddock considers the implications for countries with lower levels of disease. The answer seems to be that they aren't clear.

    The third commentary (p 998) tries to put into a clinical context the systematic reviews that show that analgesic creams and ointments may be useful in treating acute and chronic pain (p 991 and p 995). Martin R Tramèr produces some “practical clinical guidelines” that show that treatments can be useful but reflects on why it is that topical treatments are popular with patients but not doctors. The reason may, he suggests, be lack of evidence. I suspect something more primordial.

    These commentaries all, I believe, add value—not least because the authors are willing to enter the tiger country of trying to determine what evidence means. It's so hard.


    View Abstract