This site is intended for health professionals only

How much negative evidence is enough?

How many studies are required before a therapy can be said to be not effective, asks Professor Edzard Ernst

Recently, a sham-controlled randomised controlled trial by Korean acupuncturists was published testing acupuncture for smoking cessation (eCAM 2010;7:233-38). It showed that the withdrawal symptoms of individuals trying to give up smoking are not improved through real acupuncture any more than through sham-acupuncture.

Most previous studies (over one dozen) have generated similar results. A Cochrane review also failed to arrive at positive conclusions.

For me, this poses three questions:

1) How much negative evidence is enough to state with reasonable confidence that a therapy is not effective for a given condition?

2) Assuming that many clinicians earn their money by using an unproven or disproven treatment (eg acupuncturists treating smokers), we have to ask, is this dishonest?

3) Supposing that these clinicians do not know what the best evidence regarding that treatment shows, is it ethical to be ignorant?

Anyone following this blog probably can guess how I would answer these questions. I'd be interested to hear what others think.

Professor Edzard Ernst Professor Edzard Ernst