Article Text

How to read a paper. The basics of evidence based medicine

    Statistics from

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    How to read a paper. The basics of evidence based medicine. Bytrisha greenhalgh. (Pp 222, £16.95). Published by BMJ Books, London, 2001. ISBN 0 7279 1578 9

    The title of this slim paperback is somewhat misleading, as it covers considerably more than reading scientific papers (including how to deal with a visit from a drug rep!). There are two chapters which orientate the reader on why and how to keep abreast of the medical literature, including its electronic forms, nine on various aspects of reading papers, and one on implementing evidence based data. All necessary skills for the medical academic, of course, or for someone approaching a little aired but thorny issue. However, the author works in primary care, and much of her guidance seems directed towards drug trials, and particularly to the needs of her colleagues who may be wondering “how shall I treat the patient actually sitting in my surgery today?”. Good as the advice she gives may be, it is difficult to picture the general practitioner, medical registrar, or even less the tyro casualty officer, asking the patient to wait while he or she boots the computer and searches the medical literature, starting with a couple of systematic reviews and delving into an article published inRevista Médica Española, for example, only to do the same during the next consultation and, possibly, repeating the process next week, as an important new contribution may have appeared. (It is not until the penultimate chapter that we read of the existence of computerised decision support systems!)

    I apologise if my first paragraph seems somewhat tetchy because, like many hospital doctors, and particularly many long suffering radiologists, my experience has led me to appreciate only too clearly the messages Dr Greenhalgh is putting across. Maybe I was overly alienated by the almost insufferably smug image she conjures up. The Preface begins “When I wrote this book in 1996, evidence based medicine was a bit of an unknown quantity. A handful of academics (including me) were enthusiastic . . .” and on page 55 the author tells us she was a junior doctor not in any old centre, but in “a world renowned teaching hospital”. The term “evidence based medicine” may have been novel in 1996, but many of my former colleagues would, I am sure, reject the idea that the concept was new. Many more might feel miffed by her suggesting that “if you are a practising (and non-academic) physician, your main contact with published papers may well be what gets fed to you by a drug rep”.

    A book like this inevitably contains criticism of previous publications, although Dr Greenhalgh refrains from naming too many names. However, the right to be highly critical of other people's sloppy work brings with it the corresponding duty to make one's own above criticism. Medical students are among the intended targets of this book, and the literary style (“we need to hang out, listen to what people say”; “check out the literature”; researchers should “describe in detail where they are coming from”) may irritate readers more advanced in years, as may the habit of customarily according peers and professors their title(s), while using demotic forms (Dave, Nick, Andy, Sandy) for others, presumably to indicate a degree of familiarity.

    One may also quibble with certain of her ideas. She doesnot, for example, mention that one of the reasons a piece of research which is not original might be undertaken is that one simply does not believe the results in published papers, despite their apparently impeccable methodology; there are enough examples of fraudulent work in the literature for one not to be overly coy about mentioning this as a possibility. About a third of the references to the chapter entitled “Papers that tell you what things cost”, to which the author helpfully appends “(economic analyses)” are from American sources, but Dr Greenhalgh fails to make the crucial observation that most transatlantic analyses deal withcharges, notcosts, a major shortcoming which a comparison of costs of, for instance, MRI in non-profit and for-profit centres makes abundantly clear. To me, she also seems repeatedly to cop out (as she might say) when faced with rather basic philosophical questions, such as how we define health and disease and what, other than simple efficacy, can reasonably determine choices of management strategy. As a result of the first of these, she paints herself into a corner on what seems to be one of her main topics of interest, referring to the WHO definition of diabetes mellitus as the “gold standard”, so that if you conform to it “you can call yourself diabetic”, then parenthetically noting that it had changed since her first edition.

    Having got that off my chest, I must add that this book contains innumerable useful insights and thought provoking reflections (some of both original) on the biomedical literature and on research. It clearly fulfills a need, and was reprinted no fewer than six times between its original publication in 1997 and this second edition. I would guess its most appropriate audience would be medical students and young people embarking on biomedical research themselves; one would hope that they would soon pass on to the more detailed texts in the bibliography. I would be most unhappy to think that anybody could get through specialist registrar training without repeatedly having heard many of the principles expounded here from the mouths of their teachers. If they do not, they should present their seniors with a copy when they leave!