>>Even worse the situation is for treatments. There are no randomized trail regulations for treatments (not that I know off).
>
>You're kidding, right? Treatments are tested in the same way as medications, usually with double-blind studies. (In some cases, there's no ethical way to have a placebo treatment without the doctors knowing.) Insurance generally won't pay for treatments that haven't been proven by such studies.
>
>Tamar


I guess it depends what is meant by "treatments", but my impression was the same as Walter's. Kind of supported by

"In general, new surgical procedures are developed by a single surgeon or a small group of surgeons. These individuals then employ the new procedure on their patients, observe the results, and report them either as prospective or retrospective studies. This approach does not allow comparison of one procedure with others, or with a sham procedure, or (in most instances) with nonoperative treatment. Acceptance of new procedures is based on their perceived value relative to previously accepted treatments. This process can be powerfully influenced by the enthusiasm, skill, and prominence of the surgeon reporting the results and by their selection of patients for treatment. "

from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1888585/