There has been much discussion recently regarding the actual meaning of 'evidence-based' policies since they clearly don't work and are obviously not based on any genuine evidence of efficacy. Carl Phillips and Christopher Snowdon have discussed it, latterly here:
The problem with evidence-based policy
In fact it is a form of 'NGO-speak' used in public policy discussion and its meaning has no basis in the normal interpration of the words in English.
1. The name has no relation to whether it works or not. For example, in the tobacco control industry it means, "A policy agreed by senior industry figures", or "A policy that suits our agenda although it may not produce results".
As a concrete example, pharmaceutical interventions for smoking cessation such as NRTs like nicotine skin patches are agreed as 'evidence-based' despite the fact they have a 98% failure rate and cannot possibly be descibed as effective; they simply don't work and no amount of massaging the data can make it appear they work since their performance is so poor. Indeed, they are a dismal failure on all counts; not least because they are dangerous by default, since 98% of patients are therefore at a 50% risk of death by returning to smoking .
2. There is no need for such policies to work, they are designed to progress an agenda. All such policies are part of a structured process and their individual effects are irrelevant.
3. Normally, such a term is used by those who have no remit to perform the function they insist on providing for financial gain. They are neither wanted by the public nor placed in their policy-affecting position by any form of legitimate process. Therefore, half of their effort consists of attempting to legitimise their existence and the other half in pushing their agenda: commonly, to be paid by a commercial funder or funders to pursue those funders' objectives when they can be shoehorned to fit an agreed ideology that tends to be unrealistic (and perhaps unpleasant, when examined in detail).
Thus the term itself is meaningless and should not be interpreted as having some sort of scientific basis or function. It is propaganda-speak for 'a device that allows us to claim we are working for the public good though in fact we are consolidating our position, pleasing our funders, and waiting for the next stage of the political process to complete'.
For an as-yet unexplained reason, the next stage normally involves a salary increase; this is not related to any increase in benefit to the public but may be related to an increase in the propounder of the evidence-based policy's perceived credibility. Thus, credibility is a remunerable factor (though unrelated, in real terms, to any public benefit).
 All statistics were made up by clinical researchers .
 The meaning of the term 'clinical researcher', like 'evidence-based policy', is not based on standard English. It means something like, 'a person X, funded by Y, to produce result Z'.