Absence of evidence is not evidence of absence.

[Guest guest]

Stumbled across this article looking for something else.

Probably can be applied to many aspects of our clinical, financial, and personal practices.

“Absence of evidence is not evidence of absence.”

Dr. Carl Sagan quotes (American Astronomer, Writer and Scientist, 1934-1996)===========================================

www.karger.com/journals/fkm

Absence of Evidence …

An often quoted principle holds that absence of evidence (of therapeutic benefit or risk) must not be confused with evidence of absence (of benefit or risk). The principle is undoubtedly correct and important. At first sight it also seems entirely straightforward but it has been interpreted in more than one way.

Problems start when we ask, what is ‘evidence’? Some would say that personal or collective experience amounts to evidence. The above principle, however, makes little sense if anecdotes are substituted for evidence. The plural of ‘anecdote’ is ‘anecdotes’ and not ‘evidence’. The principle is valid only if evidence has the meaning that evidence-based medicine has given this term, i.e. evidence from at least one or (even better) several rigorous, randomised clinical trials.

When applied to therapeutic efficacy, we now understand the above principle to mean the following: Whenever there are no compelling trial data in favour of a given therapy x, this does not mean that this therapy does not work. Some might conclude from such a statement that it is still responsible to use or recommend treatment x because it might be efficacious. Others will argue, and I tend to agree with their view, that we require more than the mere hope of efficacy – we need positive proof to issue positive recommendations. Uncertainty about efficacy foremost means one thing: We need conclusive data which defines or refutes efficacy. In other words, it should prompt us to do the research but not normally to use or recommend therapy x for routine use.

When applied to therapeutic risk, the above-named principle is perhaps even more important: Absence of evidence of risk must not be confused with evidence of absence of risk. As long as we are unsure whether treatment x causes harm, we must not assume it to be risk-free. This looks perfectly obvious, but what are we to conclude?

Few would opt for a liberal interpretation along the following lines: As long as there is no definitive evidence of risk, it is all right to use or recommend therapy x. We simply cannot allow patients to be submitted to risks because we are uncertain.

The wisest conclusion from the above principle as applied to safety is therefore the following: Therapy x has to be regarded as unsafe until positive evidence of its safety is available.

Having dissected the ‘absence of evidence’ principle in this way, we now see that two important principles can be derived from it. Firstly, in medicine any intervention has to be categorized as inefficacious until proven otherwise. Secondly, every therapy has to be considered unsafe until there is positive proof of its safety. These are axioms which obviously have the purpose of safeguarding the position of the patient (in the same way that the axiom of jurisdiction ‘everyone is innocent until proven guilty’ protects the defendant).

It is the nature of principles that they apply across the board, i.e. for all areas of medicine. This means that the above-named principle must pertain to all therapeutic interventions. Yet one does not have to search far to find ‘pockets’ where they are at least partly suspended: psychotherapy, surgery, physical therapy and, of course, complementary/alternative medicine.

Rather than turning a blind eye to these exceptions to the rule, we should identify such areas of insufficient evidence and work towards filling the existing gaps in our knowledge – and this would be not to comply with a theoretical principle or to fulfil the needs of those working in academic ivory towers, it would mean acting in the best interests of our patients.

E. Ernst, Exeter

E. Ernst, MD, PhD, FRCP (Edin)

Department of Complementary Medicine

School of Sport and Health Sciences, University of Exeter

25 Park Road, Exeter EX2 4NT (UK)

Tel./Fax +44 13 92 42 49 89

E-mail E.Ernst@...

==========================================

Locke, MD