“…In so far as systematic reviews and meta-analysis are concerned, those involved may also be biased by wanting a particular outcome, and subtle biases can affect the selection of what to include, how to include it, and what method of meta-analysis should be used (RMITG, 1994;Clark et al, 1997;Clark, 2012). Before the RCT, the use of the Bradford Hill criteria was a way of determining causation, and among these the dose-response test can be applied to determine if a treatment may cause a particular outcome (Clark, 2011). I have suggested the first step before any controlled study is carried out to first try to determine in which types of patients treatment is successful when the same treatment is given to everyone (Clark, 2010(Clark, , 2012.…”