For clinical trials, you assume that the drug/treatment/etc DOES NOT work, and you have to prove that the drug/treatment/etc does in fact cause a response.
This is a lot harder than assuming it does work and proving it in fact does not work, similar to proofs in mathematics.
If you really want to stand up and show the world something works, get a group together (heck, sometimes you hear about in popular literature studies that only involved 10-20 people, why can't we?), EXHAUSTIVELY document your methods and procedures and use clear cut statistics and don't fudge on any of the calculations, make the original data available for anyone to analyze(while making sure to protect the participants) and people will be hard pressed to attack the study.
In the case of the liver flush, if we could find a reputable lab to do the testing, if somehow managed to get say even 10-20 people to submit their stones, and publically make available all the results, I think most doubters would be convinced, it'd be pretty hard to somehow do something shady with that many people.