For healthcare professionals only


scientific articles have hardly changed in 50 years

BMJ 2004; 328 doi: (Published 24 June 2004) Cite this as: BMJ 2004;328:1533
  1. Richard Smith, editor (rsmith{at}

    The simultaneous publication of a scientific study from 50 years ago and its current update provides an opportunity for observing changes in presentation. The over-riding impression is of little change. In the 50 years during which men have landed on the moon, computers and the internet have appeared, television and cars have been transformed, the scientific article has changed hardly at all. Does this reflect the robustness of the form or a failure of imagination? I suspect the latter.

    The 1954 article was shorter, had fewer references, slightly fewer statistical tests, more basic descriptive data, and crudely drawn figures, but the 2004 article is unusually long and resists the current temptation of statistical overkill.

    Both articles have something close to the traditional IMRaD (introduction, methods, results, and discussion) structure, but the 1954 article is more casual in mixing comments that strictly ought to be in the discussion of the results. Both papers are clearly written, but the older paper seems easier to read. In part this might be because it uses the active voice and contains slightly less jargon. The word “prospective” appears in the older paper, perhaps for the first time, and is accompanied by the largely unhelpful quote from Leigh Hunt that “He was a retrospective rather than a prospective man.” The old word for questionnaire—questionary—surprises.

    The biggest changes are in what might be called the furniture of the article. The older article has no structured abstract and no contributor, guarantor, and competing interest statements. The 2004 article includes our “what this study adds” box, one of our most popular innovations. Both papers include extensive thanks, but only the older paper gives the degrees and honours of the authors. The older paper says nothing about ethics committee approval, but the new paper tells us that there were no ethics committees in 1951. Some, I know, pine for such a time.


    • Conflict of interests RS is editor of the BMJ and accountable for all it contains.


    View Abstract