Well informed uncertainties about the effects of treatment

Editor—“Evidence based medicine” and “evidence based policy” have done much to improve decision making. However, it is refreshing to see Chalmers reject the abuse that suggests these can simply be used to identify a “best practice” that should always be followed.¹

The typical “evidence based” review identifies a large number of articles and then discards many as irrelevant and many more as methodologically unsound. The review is then based on the remaining handful and by implication all the other articles have nothing to contribute. However, detailed knowledge of the subject is often needed to judge the worth of a paper, and even methodologically flawed articles often contribute some information to the evidence base.

Typically an evidence base shows that fewer patients will benefit if given treatment A than if given treatment B. In the absence of any other information the best bet must be to choose treatment B, but nearly always other information and judgment are required as to whether the general conclusion that treatment B is better than A applies to an individual patient.

Wise doctors and policy makers will always use the evidence base to inform their decisions. But they also exercise judgment in considering all the information, formalised and informal, available to them before reaching a decision. In an uncertain world judgment free medicine (or policy making) is as bad as or worse than evidence free.