Intended for healthcare professionals

Student Education

Interpreting the evidence

BMJ 2008; 336 doi: (Published 01 January 2008) Cite this as: BMJ 2008;336:0801026
  1. Peter Davies, general practitioner principal1,
  2. Seth Jenkinson, retired general practitioner2
  1. 1Keighley Road Surgery, Halifax HX2 9LL
  2. 2Bradford

Randomised controlled trials can be approached from four viewpoints. Understanding these is vital to apply evidence in medical practice, argue Peter Davies and Seth Jenkinson

No doctor can seriously be against evidence based medicine, as described by Sackett and colleagues.1 The idea of resorting to “prejudice” or any form of non-evidence based medicine is absurd.2 And yet there are problems with evidence based medicine.34 These problems are with interpretation56 and with implementation78 of evidence. These problems are not unique to evidence based medicine; they are found in many other situations in which evidence needs interpretation.

The problems of interpretation are present at the level of doctors understanding evidence themselves.9 They are magnified when doctors try to present data about risks to patients in understandable formats.1011 The third layer of complexity comes when we try to understand how patients make sense of the information presented to them.12

We are under a duty to respect the autonomy of patients13 and to try to present the “facts” to them in a neutral unbiased way so that they can make their own well informed decisions about the likely risks and benefits of treatment.14 The difficulty in this is partly that evidence of itself does not show anything. The other difficulty is that evidence may show multiple things, and these only become apparent when considering the different ways that evidence is presented.

Ways of presenting evidence

The west of Scotland coronary prevention study (WOSCOPS) is an important study that provides evidence about the efficacy of reducing lipids for the primary prevention of coronary heart disease.15 I can present the evidence to a colleague or to a patient in one of several ways69 depending on how I want you, or a patient, to …

View Full Text

Log in

Log in through your institution


* For online subscription