Finally, in criticizing the hierarchy of evidence and the uses to which it is put in organizing evidence based care, it become necessary to ask what we would do instead. In asking this question, if we look to the past, we see a string of marginally effective or entirely ineffective, often dangerous or intrusive, time-consuming treatments
(e.g. psychosurgery, insulin coma therapy, psychoanalysis), all of which have been examined in poorly conducted research and been enthusiastically supported by health care professionals. Do we want to return to this approach organizing care? In a discussion of hierarchies of evidence, Newell and Burnard (2006) suggest the addition of two even lower levels of evidence than those in most hierarchies: opinions of our professional colleagues and evidence from our own clinical practice. They recognise that these sources of evidence are enticing (after all, they are easily available, and, it would be a strange world if you could not trust the evidence of your own eyes or those of people whom you know and trust). At the same time, they question the wisdom of trusting such evidence more than the work of, for example, a review which requires ongoing commitment over a period of months or years. As clinicians, we have a vested interest in believing we are successful with clients, and this is likely to colour our judgement of the interventions we use. The checks and balances inherent in the review procedures embodied in hierarchies of evidence provide a potent antidote to this potential to adopt a blinkered view of our treatments.