Editor's choice

The agonies of evidence

John Maynard Keynes argued that politicians don't like evidence. It ties their hands. I once quoted this when introducing a government minister, who was opening a conference on research. "Balls," she answered. But this week's BMJ illustrates how evidence can complicate the lives of politicians.

The arrival of a Cochrane review suggesting that mammography did not save lives but simply increased mastectomies must have been painful for those who have invested heavily in screening programmes (27 October, p 956. The British government doesn't accept the results of the Cochrane review and has stated that screening does reduce mortality (p 1088). Meanwhile, Peter Gotzsche, an author of the Cochrane review (and a member of the BMJ 's editorial board), criticises the NHS cancer screening programme for inaccurate, unscientific, and anonymous criticisms of the review (p 1131). The BMJ will return to this subject and hope to clarify it for women and doctors alike.

Evidence is also creating difficulties for politicians in relation to the National Institute for Clinical Excellence (NICE). It has decided that new treatments for multiple sclerosis are not cost effective and should not be used in the NHS (p 1087). Patients and neurologists don't like this decision, so the government has tried to compromise by setting up a large trial of the new treatments. NICE is also to make a decision on imatinib, a new treatment for chronic myeloid leukaemia but the decision is not expected until next yeargiving us the new phrase "NICE blight." The manufacturers have been providing the drug for free, but now the Department of Health is telling health authorities to pay for it. Again, NICE is being circumvented.

Politicians will also have a problem with evidence showing that intensive case management of patients with severe mental illness does not reduce violence any more than standard care (p 1093). Intensive case management has been a "solution" in various places to the problem of psychotic patients (rarely) being violent.

Politicians can, of course, always find a way to ignore inconvenient evidence: more research is needed; other data lead to contrary conclusions; or patient experience makes us sceptical of the results. Doctors also put evidence on one side. Two researchers from the west of England have used qualitative research to understand why general practitioners don't implement evidence (p 1100). Unsurprisingly, professional and personal experience is important. One doctor reaches for a "decent dose of warfarin and digoxin no hesitation at all" to treat patients with atrial fibrillation because of the death of a grandfather. Another doctor is reluctant because of the loss of a patient on warfarin.

Footnotes

To receive Editor's choice by email each week subscribe via our website: www.bmj.com/cgi/customalert