Results
Our findings suggest that when the difference between novel treatments and the historical control is at least one logarithm (base 10) of magnitude, the veracity of testing in non-RCTs seems to be established.